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.
167 JH/36 Have the general router option retry_use_local_part default to true when
168 any of the restrictive preconditions are set (to anything). Previously it
169 was only for check_local user. The change removes one item of manual
170 configuration which is required for proper retries when a remote router
171 handles a subset of addresses for a domain.
173 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
174 link count into consideration.
180 JH/01 Remove code calling the customisable local_scan function, unless a new
181 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
183 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
184 non-signal-safe functions being used.
186 JH/03 Bug 2269: When presented with a received message having a stupidly large
187 number of DKIM-Signature headers, disable DKIM verification to avoid
188 a resource-consumption attack. The limit is set at twenty.
190 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
191 report of oldest_pass in ${authres } in consequence, and separate out
192 some descriptions of reasons for verification fail.
194 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
195 files in the spool were present and unlocked. A queue-runner could spot
196 them, resulting in a duplicate delivery. Fix that by doing the unlock
197 after the unlink. Investigation by Tim Stewart. Take the opportunity to
198 add more error-checking on spoolfile handling while that code is being
201 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
202 No known attacks, no CVE, this is defensive hardening.
204 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
205 a queue-runner could start a delivery while other operations were ongoing.
206 Cutthrough delivery was a common victim, resulting in duplicate delivery.
207 Found and investigated by Tim Stewart. Fix by using the open message data
208 file handle rather than opening another, and not locally closing it (which
209 releases a lock) for that case, while creating the temporary .eml format
210 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
212 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
213 $sender_verify_failure/$recipient_verify_failure to "random".
215 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
218 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
219 Previously this would segfault.
221 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
224 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
225 like zero, since the resolver should be doing this for us, But we need one
226 as a CNAME but no MX presence gets the CNAME returned; we need to check
227 that doesn't point to an MX to declare it "no MX returned" rather than
228 "error, loop". A new main option is added so the older capability of
229 following some limited number of chain links is maintained.
231 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
233 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
234 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
235 now-deprecated earlier definitions used only specified the range up to TLS
236 1.2 (in the older-version library docs).
238 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
240 JH/15 Rework TLS client-side context management. Stop using a global, and
241 explicitly pass a context around. This enables future use of TLS for
242 connections to service-daemons (eg. malware scanning) while a client smtp
243 connection is using TLS; with cutthrough connections this is quite likely.
245 JH/16 Fix ARC verification to do AS checks in reverse order.
247 JH/17 Support a "tls" option on the ${readsocket } expansion item.
249 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
250 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
251 Previously the "utf8" would be re-prepended for every additional message.
253 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
254 Previously thery were accepted, resulting in issues when attempting to
255 forward messages to a non-supporting MTA.
257 PP/02 Let -n work with printing macros too, not just options.
259 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
260 one parent address was copied, and bogus data was used at delivery-logging
261 time. Either a crash (after delivery) or bogus log data could result.
262 Discovery and analysis by Tim Stewart.
264 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
265 Previously if the string ended mid-character, we did not insert the
266 promised '?' replacement.
268 PP/04 Documentation: current string operators work on bytes, not codepoints.
270 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
271 should pack well giving a smaller memory footprint so better caching and
272 therefore performance. Group the declarations where this can't be done so
273 that the byte-sized flag variables are not interspersed among pointer
274 variables, giving a better chance of good packing by the compiler.
276 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
277 non-null, to avoid issues with sites running BATV. Previously reports were
278 sent with an empty envelope sender so looked like bounces.
280 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
281 The ignore_error flag wasn't being returned from the filter subprocess so
282 was not set for later routers. Investigation and fix by Matthias Kurz.
284 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
285 and a msg:complete for the whole, when a message is manually removed using
286 -Mrm. Developement by Matthias Kurz, hacked on by JH.
288 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
289 a "Gnu special" function, asprintf() in the DB utility binary builds; I
290 hope that is portable enough.
292 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
293 requiring a known-CA anchor certificate; make it now rely entirely on the
294 TLSA as an anchor. Checking the name on the leaf cert against the name
295 on the A-record for the host is still done for TA (but not for EE mode).
297 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
298 included in delivery lines for non-proxied connections, when compiled with
299 SUPPORT_SOCKS and running with proxy logging enabled.
301 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
302 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
303 move the existing event to fire before the normal logging of message
304 failure so that custom logging is bracketed by normal logging.
306 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
307 msg:fail:internal event. Developement by Matthias Kurz.
309 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
310 far too small for todays use of crypto signatures stored there. Go all
311 the way to the max DNS message size of 64kB, even though this might be
312 overmuch for IOT constrained device use.
314 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
315 copy a string over itself. The library routine is documented as not
316 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
318 JH/32 For main options check_spool_space and check_inode_space, where the
319 platform supports 64b integers, support more than the previous 2^31 kB
320 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
321 the previous G, M, k.
323 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
324 $authenticated_fail_id variable on authentication failure. Previously
327 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
328 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
329 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
330 GNUTLS_SEC_PARAM_MEDIUM.
332 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
333 side. Previously we would continue as if no SNI had been received.
335 JH/36 Harden the handling of string-lists. When a list consisted of a sole
336 "<" character, which should be a list-separator specification, we walked
337 off past the nul-terimation.
339 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
340 causes) even when the retry time is not yet met. Previously they were
341 not, meaning that when (say) an account was over-quota and temp-rejecting,
342 and multiple senders' messages were queued, only one sender would get
343 notified on each configured delay_warning cycle.
345 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
347 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
348 cipher-suites, an error can be left on the stack even for a succeeding
349 accept; this results in impossible error messages when a later operation
352 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
353 return error codes indicating retry. Under TLS1.3 this becomes required.
355 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
356 it only wrote the new authenticators, resulting in a lack of tracking of
357 peer changes of ESMTP extensions until the next cache flush.
359 JH/41 Fix the loop reading a message header line to check for integer overflow,
360 and more-often against header_maxsize. Previously a crafted message could
361 induce a crash of the recive process; now the message is cleanly rejected.
363 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
364 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
370 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
371 When redis_servers is set to a list of > 1 element, and the Redis servers
372 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
373 case of MOVED into a DEFER case instead, thus moving the query onto the
374 next server in the list. For a cluster of N elements, all N servers must
375 be defined in redis_servers.
377 GF/02 Catch and remove uninitialized value warning in exiqsumm
378 Check for existence of @ARGV before looking at $ARGV[0]
380 JH/01 Replace the store_release() internal interface with store_newblock(),
381 which internalises the check required to safely use the old one, plus
382 the allocate and data copy operations duplicated in both (!) of the
383 extant use locations.
385 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
386 modifier. This matches the restriction on the commandline.
388 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
389 Previously only the last row was returned.
391 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
392 we assumed that tags in the header were well-formed, and parsed the
393 element content after inspecting only the first char of the tag.
394 Assumptions at that stage could crash the receive process on malformed
397 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
398 While running the DKIM ACL we operate on the Permanent memory pool so that
399 variables created with "set" persist to the DATA ACL. Also (at any time)
400 DNS lookups that fail create cache records using the Permanent pool. But
401 expansions release any allocations made on the current pool - so a dnsdb
402 lookup expansion done in the DKIM ACL releases the memory used for the
403 DNS negative-cache, and bad things result. Solution is to switch to the
404 Main pool for expansions.
405 While we're in that code, add checks on the DNS cache during store_reset,
406 active in the testsuite.
407 Problem spotted, and debugging aided, by Wolfgang Breyha.
409 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
410 When none of the hosts presented to a transport match an already-open
411 connection, close it and proceed with the list. Previously we would
412 queue the message. Spotted by Lena with Yahoo, probably involving
415 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
416 Previously a spurious "250 OK id=" response was appended to the proper
419 JH/08 The "support for" informational output now, which built with Content
420 Scanning support, has a line for the malware scanner interfaces compiled
421 in. Interface can be individually included or not at build time.
423 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
424 by the template makefile "src/EDITME". The "STREAM" support for an older
425 ClamAV interface method is removed.
427 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
428 rows affected is given instead).
430 JH/11 The runtime Berkeley DB library version is now additionally output by
431 "exim -d -bV". Previously only the compile-time version was shown.
433 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
434 SMTP connection. Previously, when one had more receipients than the
435 first, an abortive onward connection was made. Move to full support for
436 multiple onward connections in sequence, handling cutthrough connection
437 for all multi-message initiating connections.
439 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
440 routers. Previously, a multi-recipient message would fail to match the
441 onward-connection opened for the first recipient, and cause its closure.
443 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
444 a timeout on read on a GnuTLS initiating connection, resulting in the
445 initiating connection being dropped. This mattered most when the callout
446 was marked defer_ok. Fix to keep the two timeout-detection methods
449 JH/15 Relax results from ACL control request to enable cutthrough, in
450 unsupported situations, from error to silently (except under debug)
451 ignoring. This covers use with PRDR, frozen messages, queue-only and
454 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
456 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
457 metadata, resulting in a crash in free().
459 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
460 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
461 Broken also in d185889f4, with init system revamp.
463 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
464 Previously we abruptly closed the connection after reading a malware-
465 found indication; now we go on to read the "scan ok" response line,
468 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
469 ACL. Previously, a crash would result.
471 JH/19 Speed up macro lookups during configuration file read, by skipping non-
472 macro text after a replacement (previously it was only once per line) and
473 by skipping builtin macros when searching for an uppercase lead character.
475 JH/20 DANE support moved from Experimental to mainline. The Makefile control
476 for the build is renamed.
478 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
479 was allocated for every new TLS startup, meaning one per message. Fix
480 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
482 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
483 reported the original. Fix to report (as far as possible) the ACL
484 result replacing the original.
486 JH/23 Fix memory leak during multi-message connections using STARTTLS under
487 OpenSSL. Certificate information is loaded for every new TLS startup,
488 and the resources needed to be freed.
490 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
492 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
495 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
496 DATA response info to the (existing) per-recipient response info for
497 the "C=" log element. It can have useful tracking info from the
498 destination system. Patch from Simon Arlott.
500 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
501 length value. Previously this would segfault.
503 HS/02 Support Avast multiline protoocol, this allows passing flags to
504 newer versions of the scanner.
506 JH/28 Ensure that variables possibly set during message acceptance are marked
507 dead before release of memory in the daemon loop. This stops complaints
508 about them when the debug_store option is enabled. Discovered specifically
509 for sender_rate_period, but applies to a whole set of variables.
510 Do the same for the queue-runner and queue-list loops, for variables set
511 from spool message files. Do the same for the SMTP per-message loop, for
512 certain variables indirectly set in ACL operations.
514 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
515 as a multi-recipient message from a mailinglist manager). The coding had
516 an arbitrary cutoff number of characters while checking for more input;
517 enforced by writing a NUL into the buffer. This corrupted long / fast
518 input. The problem was exposed more widely when more pipelineing of SMTP
519 responses was introduced, and one Exim system was feeding another.
520 The symptom is log complaints of SMTP syntax error (NUL chars) on the
521 receiving system, and refused recipients seen by the sending system
522 (propating to people being dropped from mailing lists).
523 Discovered and pinpointed by David Carter.
525 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
526 replaced by the ${authresults } expansion.
528 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
530 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
531 allows proper process termination in container environments.
533 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
534 Previously the "final dot" had a newline after it; ensure it is CR,LF.
536 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
537 and "err_perm", deprecated since 4.83 when the RFC-defined words
538 "temperror" and "permerror" were introduced.
540 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
541 transport-filters or DKIM-signing. The restriction was lost in the
542 consolidation of verify-callout and delivery SMTP handling.
543 Extend the restriction to also cover ARC-signing.
545 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
546 in defer=pass mode supply a 450 to the initiator. Previously the message
549 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
550 tls_require_ciphers is used as before.
552 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
553 "pass_unscanned". Only tmpfails from the scanner are written to
554 the paniclog, as they may require admin intervention (permission
555 denied, license issues). Other scanner errors (like decompression
556 bombs) do not cause a paniclog entry.
558 JH/36 Fix reinitialisation of DKIM logging variable between messages.
559 Previously it was possible to log spurious information in receive log
562 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
563 triggered odd behaviour from Outlook Express clients.
565 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
568 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
569 since the IETF WG has not yet settled on that versus the original
570 "bare" representation.
572 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
573 Previously the millisecond value corrupted the output.
574 Fix also for syslog_pid=no and log_selector +pid, for which the pid
575 corrupted the output.
581 JH/01 Rework error string handling in TLS interface so that the caller in
582 more cases is responsible for logging. This permits library-sourced
583 string to be attached to addresses during delivery, and collapses
584 pairs of long lines into single ones.
586 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
587 during configuration. Wildcards are allowed and expanded.
589 JH/02 Rework error string handling in DKIM to pass more info back to callers.
590 This permits better logging.
592 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
593 do not close it down and have the child transport start it up again on
594 the passed-on TCP connection. Instead, proxy the child (and any
595 subsequent ones) for TLS via a unix-domain socket channel. Logging is
596 affected: the continued delivery log lines do not have any DNSSEC, TLS
597 Certificate or OCSP information. TLS cipher information is still logged.
599 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
600 identical IP addresses on different listening ports. Will also affect
603 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
604 add noisy ifdef guards to special-case this sillyness.
605 Patch from Bernd Kuhls.
607 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
608 than 255 are no longer allowed.
610 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
611 Disable the session-cache too, which might reduce our load. Since we
612 currrectly use a new context for every connection, both as server and
613 client, there is no benefit for these.
614 GnuTLS appears to not support tickets server-side by default (we don't
615 call gnutls_session_ticket_enable_server()) but client side is enabled
616 by default on recent versions (3.1.3 +) unless the PFS priority string
619 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
620 <https://reproducible-builds.org/specs/source-date-epoch/>.
622 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
623 the check for any unsuccessful recipients did not notice the limit, and
624 erroneously found still-pending ones.
626 JH/08 Pipeline CHUNKING command and data together, on kernels that support
627 MSG_MORE. Only in-clear (not on TLS connections).
629 JH/09 Avoid using a temporary file during transport using dkim. Unless a
630 transport-filter is involved we can buffer the headers in memory for
631 creating the signature, and read the spool data file once for the
632 signature and again for transmission.
634 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
635 4.77 as the kernel support then wasn't solid, having issues in 64bit
636 mode. Now, it's been long enough. Add support for FreeBSD also.
638 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
639 case where the routing stage had gathered several addresses to send to
640 a host before calling the transport for the first, we previously failed
641 to close down TLS in the old transport process before passing the TCP
642 connection to the new process. The new one sent a STARTTLS command
643 which naturally failed, giving a failed delivery and bloating the retry
644 database. Investigation and fix prototype from Wolfgang Breyha.
646 JH/12 Fix check on SMTP command input synchronisation. Previously there were
647 false-negatives in the check that the sender had not preempted a response
648 or prompt from Exim (running as a server), due to that code's lack of
649 awareness of the SMTP input buffering.
651 PP/04 Add commandline_checks_require_admin option.
652 Exim drops privileges sanely, various checks such as -be aren't a
653 security problem, as long as you trust local users with access to their
654 own account. When invoked by services which pass untrusted data to
655 Exim, this might be an issue. Set this option in main configuration
656 AND make fixes to the calling application, such as using `--` to stop
659 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
660 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
661 responses to those, into a single TLS record each way (this usually means
662 a single packet). As a side issue, smtp_enforce_sync now works on TLS
665 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
666 affects you only if you're dancing at the edge of the param size limits.
667 If you are, and this message makes sense to you, then: raise the
668 configured limit or use OpenSSL 1.1. Nothing we can do for older
671 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
672 cmdline element to get the documented default one. Previously it was
675 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
676 are unsuable for non-message-specific reasons - eg. connection timeout,
677 banner-time rejection.
679 JH/16 Fix logging of delivery remote port, when specified by router, under
682 PP/06 Repair manualroute's ability to take options in any order, even if one
683 is the name of a transport.
686 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
688 JH/17 Change the list-building routines interface to use the expanding-string
689 triplet model, for better allocation and copying behaviour.
691 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
692 Previously it was constructed the first time a possibly-matching string
693 was met in the configuration file input during startup; now it is done
696 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
697 compatible one, to avoid the (poorly documented) possibility of a config
698 file in the working directory redirecting the DB files, possibly correpting
699 some existing file. CVE-2017-10140 assigned for BDB.
701 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
702 cache-hot. Previously, although the result was properly cached, the
703 initial verify call returned a defer.
705 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
706 the main verify for receipient in uncached-mode.
708 JH/22 Retire historical build files to an "unsupported" subdir. These are
709 defined as "ones for which we have no current evidence of testing".
711 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
712 if present. Previously it was ignored.
714 JH/24 Start using specified-initialisers in C structure init coding. This is
715 a C99 feature (it's 2017, so now considered safe).
717 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
718 if was a fixed-sized field and bitmask ops via macros; it is now more
721 PP/07 GitHub PR 56: Apply MariaDB build fix.
722 Patch provided by Jaroslav Škarvada.
724 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
725 during Coverity cleanups [4.87 JH/47]
726 Diagnosis and fix provided by Michael Fischer v. Mollard.
728 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
729 the right size to place the terminating semicolon on its own folded
730 line, the header hash was calculated to an incorrect value thanks to
731 the (relaxed) space the fold became.
733 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
734 and confused the parent.
736 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
737 which could crash as a result. This could lead to undeliverable messages.
739 JH/28 Logging: "next input sent too soon" now shows where input was truncated
742 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
743 matters on fast-turnover and PID-randomising systems, which were getting
744 out-of-order delivery.
746 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
747 a possibly-overlapping copy. The symptom was that "Remote host closed
748 connection in response to HELO" was logged instead of the actual 4xx
751 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
752 Previously only that bufferd was discarded, resulting in SYMTP command
755 JH/32 DKIM: when a message has multiple signatures matching an identity given
756 in dkim_verify_signers, run the dkim acl once for each. Previously only
757 one run was done. Bug 2189.
759 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
760 "panic the current process" to "deliberately defer". The panic log is
761 still written with the problem list name; the mail and reject logs now
762 get a temp-reject line for the message that was being handled, saying
763 something like "domains check lookup or other defer". The SMTP 451
764 message is still "Temporary local problem".
766 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
767 A crafted sequence of BDAT commands could result in in-use memory beeing
768 freed. CVE-2017-16943.
770 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
771 from SMTP input. Previously it was always done; now only done for DATA
772 and not BDAT commands. CVE-2017-16944.
774 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
775 to the message (such as an overlong header line). Previously this was
776 not done and we did not exit BDAT mode. Followon from the previous item
777 though a different problem.
783 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
784 than -2003 did; needs libidn2 in addition to libidn.
786 JH/02 The path option on a pipe transport is now expanded before use.
788 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
789 Patch provided by "Björn", documentation fix added too.
791 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
792 missing a wire-to-host endian conversion.
794 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
795 close after a BDAT command line could be taken as a following command,
796 giving a synch failure. Fix by only checking for synch immediately
797 before acknowledging the chunk.
799 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
800 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
801 macro. Patches provided by Josh Soref.
803 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
804 Previously we did not; the RFC seems ambiguous and VRFY is not listed
805 by IANA as a service extension. However, John Klensin suggests that we
808 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
809 the dkim code may be unix-mode line endings rather than smtp wire-format
810 CRLF, so prepend a CR to any bare LF.
812 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
813 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
815 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
816 the first were themselves being wrongly included in the feed into dkim
817 processing; with most chunk sizes in use this resulted in an incorrect
818 body hash calculated value.
820 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
821 DKIM signature block, for verification. Although advised against by
822 standards it is specifically not ruled illegal.
824 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
826 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
827 missing a body hash (the bh= tag).
829 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
830 It seems that HAProxy sends the Proxy Protocol information in clear and
831 only then does a TLS startup, so do the same.
833 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
834 TCP connections (such as for Spamd) unless the daemon successfully set
835 Fast Open mode on its listening sockets. This fixes breakage seen on
836 too-old kernels or those not configured for Fast Open, at the cost of
837 requiring both directions being enabled for TFO, and TFO never being used
838 by non-daemon-related Exim processes.
840 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
841 endings, at least on the first header line. Try to canonify any that get
842 past that check, despite the cost.
844 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
845 now limited to an arbitrary five deep, while parsing addresses with the
846 strip_excess_angle_brackets option enabled.
848 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
849 instead leave the unprompted TLS handshake in socket buffer for the
850 TLS library to consume.
852 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
854 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
856 JH/16 Drop variables when they go out of scope. Memory management drops a whole
857 region in one operation, for speed, and this leaves assigned pointers
858 dangling. Add checks run only under the testsuite which checks all
859 variables at a store-reset and panics on a dangling pointer; add code
860 explicitly nulling out all the variables discovered. Fixes one known
861 bug: a transport crash, where a dangling pointer for $sending_ip_address
862 originally assigned in a verify callout, is re-used.
864 PP/06 Drop '.' from @INC in various Perl scripts.
866 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
868 PP/08 Reduce a number of compilation warnings under clang; building with
869 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
870 should be warning-free.
872 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
874 HS/01 Fix portability problems introduced by PP/08 for platforms where
875 realloc(NULL) is not equivalent to malloc() [SunOS et al].
877 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
878 chunk. This allows us to accept broken chunked messages. We need a more
879 general solution here.
881 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
882 already-broken messages in the queue.
884 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
886 JH/19 Fix reference counting bug in routing-generated-address tracking.
892 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
893 supports it and a size is available (ie. the sending peer gave us one).
895 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
896 years of being deprecated). The replacements are the ACLs
897 acl_smtp_mime and acl_not_smtp_mime.
899 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
900 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
901 or even an in-clear connection were permitted. Now, if the host lookup
902 was dnssec and dane was requested then the host is only used if the
903 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
904 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
905 if one fails this test.
906 This means that a poorly-configured remote DNS will make it incommunicado;
907 but it protects against a DNS-interception attack on it.
909 JH/04 Bug 1810: make continued-use of an open smtp transport connection
910 non-noisy when a race steals the message being considered.
912 JH/05 If main configuration option tls_certificate is unset, generate a
913 self-signed certificate for inbound TLS connections.
915 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
916 in rewrites and routers.
918 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
919 and logged a warning sing 4.83; now they are a configuration file error.
921 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
922 (lacking @domain). Apply the same qualification processing as RCPT.
924 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
926 JH/10 Support ${sha256:} applied to a string (as well as the previous
929 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
930 a cutthrough deliver is pending, as we always want to make a connection.
931 This also avoids re-routing the message when later placing the cutthrough
932 connection after a verify cache hit.
933 Do not update it with the verify result either.
935 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
936 when routing results in more than one destination address.
938 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
939 signing (which inhibits the cutthrough capability). Previously only
940 the presence of an option was tested; now an expansion evaluating as
941 empty is permissible (obviously it should depend only on data available
942 when the cutthrough connection is made).
944 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
945 the relevant preceding SMTP command did not note the pipelining mode.
947 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
948 Previously they were not counted.
950 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
951 as one having no matching records. Previously we deferred the message
952 that needed the lookup.
954 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
955 distinguished as "(=".
957 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
958 for missing MX records. Previously it only worked for missing A records.
960 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
962 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
963 after the data-go-ahead and data-ack. Patch from Jason Betts.
965 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
966 even for a "none" policy. Patch from Tony Meyer.
968 JH/22 Fix continued use of a connection for further deliveries. If a port was
969 specified by a router, it must also match for the delivery to be
972 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
973 When one of the recipients of a message was unsuitable for the connection
974 (has no matching addresses), we lost track of needing to mark it
975 deferred. As a result mail would be lost.
977 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
979 JH/25 Decoding ACL controls is now done using a binary search; the source code
980 takes up less space and should be simpler to maintain. Merge the ACL
981 condition decode tables also, with similar effect.
983 JH/26 Fix problem with one_time used on a redirect router which returned the
984 parent address unchanged. A retry would see the parent address marked as
985 delivered, so not attempt the (identical) child. As a result mail would
988 JH/27 Fix a possible security hole, wherein a process operating with the Exim
989 UID can gain a root shell. Credit to http://www.halfdog.net/ for
990 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
993 JH/28 Enable {spool,log} filesystem space and inode checks as default.
994 Main config options check_{log,spool}_{inodes,space} are now
995 100 inodes, 10MB unless set otherwise in the configuration.
997 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
998 Previously it only applied to the main-section connection policy
1001 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1003 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1004 by me. Added RFC7919 DH primes as an alternative.
1006 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1007 are not in the system include path.
1009 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1010 GnuTLS, when a session startup failed (eg because the client disconnected)
1011 Exim did stdio operations after fclose. This was exposed by a recent
1012 change which nulled out the file handle after the fclose.
1014 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1015 signed directly by the cert-signing cert, rather than an intermediate
1016 OCSP-signing cert. This is the model used by LetsEncrypt.
1018 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1020 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1021 an incoming connection.
1023 HS/02 Bug 1802: Do not half-close the connection after sending a request
1026 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1027 fallback to "prime256v1".
1029 JH/34 SECURITY: Use proper copy of DATA command in error message.
1030 Could leak key material. Remotely exploitable. CVE-2016-9963.
1036 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1037 and 3.4.4 - once the server is enabled to respond to an OCSP request
1038 it does even when not requested, resulting in a stapling non-aware
1039 client dropping the TLS connection.
1041 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1042 support variable-length bit vectors. No functional change.
1044 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1045 The I= interface field on outgoing lines is now after the H= remote
1046 host field, same as incoming lines. There is a separate
1047 outgoing_interface log selector which allows you to disable the
1050 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1051 If not running log_selector +smtp_connection the mainlog would be held
1052 open indefinitely after a "too many connections" event, including to a
1053 deleted file after a log rotate. Leave the per net connection logging
1054 leaving it open for efficiency as that will be quickly detected by the
1055 check on the next write.
1057 HS/01 Bug 1671: Fix post transport crash.
1058 Processing the wait-<transport> messages could crash the delivery
1059 process if the message IDs didn't exist for some reason. When
1060 using 'split_spool_directory=yes' the construction of the spool
1061 file name failed already, exposing the same netto behaviour.
1063 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1064 mime_regex ACL conditions.
1066 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1067 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1068 to HELO, local diagnostic string.
1070 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1071 log line to debug. Even when configured with a tls authenticator many
1072 client connections are expected to not authenticate in this way, so
1073 an authenticate fail is not an error.
1075 HS/02 Add the Exim version string to the process info. This way exiwhat
1076 gives some more detail about the running daemon.
1078 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1079 matter for fast-change records such as DNSBLs.
1081 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1082 retry record, even if constant. There may be multiple transports with
1083 different interface settings and the retry behaviour needs to be kept
1086 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1088 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1090 JH/10 Bug 840: fix log_defer_output option of pipe transport
1092 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1093 hosts_randomize. This matters a lot when combined with mua_wrapper.
1095 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1096 ${quote_pgsql:<string>} operator.
1098 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1100 JH/14 Change header file name for freeradius-client. Relevant if compiling
1101 with Radius support; from the Gentoo tree and checked under Fedora.
1103 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1105 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1106 option transport_filter, meaning no filtering.
1108 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1110 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1111 defaults to "*" (all hosts). The variable is now available when not built
1112 with TLS, default unset, mainly to enable keeping the testsuite sane.
1113 If a server certificate is not supplied (via tls_certificate) an error is
1114 logged, and clients will find TLS connections fail on startup. Presumably
1115 they will retry in-clear.
1116 Packagers of Exim are strongly encouraged to create a server certificate
1117 at installation time.
1119 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1120 with the $config_file variable.
1122 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1123 in transport context, after the attempt, and per-recipient. The latter type
1124 is per host attempted. The event data is the error message, and the errno
1125 information encodes the lookup type (A vs. MX) used for the (first) host,
1126 and the trailing two digits of the smtp 4xx response.
1128 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1129 to write to mainlog (or rejectlog, paniclog) in the window between file
1130 creation and permissions/ownership being changed. Particularly affects
1131 installations where exicyclog is run as root, rather than exim user;
1132 result is that the running daemon panics and dies.
1134 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1136 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1137 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1138 "pri" and "weight". Note that the previous implicit priority given by the
1139 list order is no longer honoured.
1141 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1142 for DKIM processing.
1144 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1145 by defining SUPPORT_SOCKS.
1147 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1148 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1149 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1150 variables are renamed to proxy_{local,external}_{address,port}.
1152 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1153 for a build by defining SUPPORT_I18N
1155 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1156 of the query string, and make ${quote_redis:} do that quoting.
1158 JH/29 Move Events support from Experimental to mainline, enabled by default
1159 and removable for a build by defining DISABLE_EVENT.
1161 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1163 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1164 cached by the daemon.
1166 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1167 by defining LOOKUP_REDIS. The libhiredis library is required.
1169 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1170 keys are given for lookup.
1172 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1173 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1174 only supported when built with TLS support. The PolarSSL SHA routines
1175 are still used when the TLS library is too old for convenient support.
1177 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1178 openssl_options), for security. OpenSSL forces this from version 1.1.0
1179 server-side so match that on older versions.
1181 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1182 allocation for $value could be released as the expansion processing
1183 concluded, but leaving the global pointer active for it.
1185 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1186 and to use the domains and local_parts ACL conditions.
1188 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1189 incorrectly not doubled on cutthrough transmission, hence seen as a
1190 body-termination at the receiving system - resulting in truncated mails.
1191 Commonly the sender saw a TCP-level error, and retransmitted the message
1192 via the normal store-and-forward channel. This could result in duplicates
1193 received - but deduplicating mailstores were liable to retain only the
1194 initial truncated version.
1196 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1198 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1200 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1201 we're in there, support oversigning also; bug 1309.
1203 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1205 HS/04 Add support for keep_environment and add_environment options.
1207 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1208 either intentional arithmetic overflow during PRNG, or testing config-
1211 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1212 delivery resulted in actual delivery. Cancel cutthrough before DATA
1215 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1216 rejects a recipient: pass the reject to the originator.
1218 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1219 Many were false-positives and ignorable, but it's worth fixing the
1222 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1223 for the new environment-manipulation done at startup. Move the routines
1224 from being local to tls.c to being global via the os.c file.
1226 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1227 an extract embedded as result-arg for a map, the first arg for extract
1228 is unavailable so we cannot tell if this is a numbered or keyed
1229 extraction. Accept either.
1235 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1238 JH/02 The smtp transport option "multi_domain" is now expanded.
1240 JH/03 The smtp transport now requests PRDR by default, if the server offers
1243 JH/04 Certificate name checking on server certificates, when exim is a client,
1244 is now done by default. The transport option tls_verify_cert_hostnames
1245 can be used to disable this per-host. The build option
1246 EXPERIMENTAL_CERTNAMES is withdrawn.
1248 JH/05 The value of the tls_verify_certificates smtp transport and main options
1249 default to the word "system" to access the system default CA bundle.
1250 For GnuTLS, only version 3.0.20 or later.
1252 JH/06 Verification of the server certificate for a TLS connection is now tried
1253 (but not required) by default. The verification status is now logged by
1254 default, for both outbound TLS and client-certificate supplying inbound
1257 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1260 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1261 Status Notification (bounce) messages are now MIME format per RFC 3464.
1262 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1263 under the control of the dsn_advertise_hosts option, and routers may
1264 have a dsn_lasthop option.
1266 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1267 default, modifiable by a malware= option. The list separator for
1268 the options can now be changed in the usual way. Bug 68.
1270 JH/10 The smtp_receive_timeout main option is now expanded before use.
1272 JH/11 The incoming_interface log option now also enables logging of the
1273 local interface on delivery outgoing connections.
1275 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1276 if the interface and destination host and port all match.
1278 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1281 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1282 Patch from Andrew Lewis.
1284 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1285 now supports optional time-restrictions, weighting, and priority
1286 modifiers per server. Patch originally by <rommer@active.by>.
1288 JH/16 The spamd_address main option now supports a mixed list of local
1289 and remote servers. Remote servers can be IPv6 addresses, and
1290 specify a port-range.
1292 JH/17 Bug 68: The spamd_address main option now supports an optional
1293 timeout value per server.
1295 JH/18 Bug 1581: Router and transport options headers_add/remove can
1296 now have the list separator specified.
1298 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1301 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1304 JH/21 Support for the A6 type of dns record is withdrawn.
1306 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1307 rather than the verbs used.
1309 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1310 from 255 to 1024 chars.
1312 JH/24 Verification callouts now attempt to use TLS by default.
1314 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1315 are generic router options now. The defaults didn't change.
1317 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1318 Original patch from Alexander Shikoff, worked over by JH.
1320 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1321 files if ZCAT_COMMAND is not executable.
1323 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1325 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1327 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1328 Normally benign, it bites when the pair was led to by a CNAME;
1329 modern usage is to not canonicalize the domain to a CNAME target
1330 (and we were inconsistent anyway for A-only vs AAAA+A).
1332 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1334 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1335 when evaluating $sender_host_dnssec.
1337 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1338 $sender_helo_dnssec variable.
1340 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1342 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1344 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1346 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1347 documented as working, but never had. Support all but $spam_report.
1349 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1350 added for tls authenticator.
1352 HS/03 Add perl_taintmode main config option
1358 TL/01 When running the test suite, the README says that variables such as
1359 no_msglog_check are global and can be placed anywhere in a specific
1360 test's script, however it was observed that placement needed to be near
1361 the beginning for it to behave that way. Changed the runtest perl
1362 script to read through the entire script once to detect and set these
1363 variables, reset to the beginning of the script, and then run through
1364 the script parsing/test process like normal.
1366 TL/02 The BSD's have an arc4random API. One of the functions to induce
1367 adding randomness was arc4random_stir(), but it has been removed in
1368 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1369 function when detected.
1371 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1372 cause callback expansion.
1374 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1375 syntax errors in an expansion can be treated as a string instead of
1376 logging or causing an error, due to the internal use of bool_lax
1377 instead of bool when processing it.
1379 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1380 server certificates when making smtp deliveries.
1382 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1384 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1386 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1388 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1389 Merged patch from Sebastian Wiedenroth.
1391 JH/05 Fix results-pipe from transport process. Several recipients, combined
1392 with certificate use, exposed issues where response data items split
1393 over buffer boundaries were not parsed properly. This eventually
1394 resulted in duplicates being sent. This issue only became common enough
1395 to notice due to the introduction of connection certificate information,
1396 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1398 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1399 size buffer was used, resulting in syntax errors when an expansion
1402 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1403 version 3.3.6 or later.
1405 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1406 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1407 both become "event_action", the variables become $event_name, $event_data
1408 and $event_defer_errno. There is a new variable $verify_mode, usable in
1409 routers, transports and related events. The tls:cert event is now also
1410 raised for inbound connections, if the main configuration event_action
1413 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1414 early OCSP support, but no stapling (appears to be less than 1.0.0).
1416 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1417 server certificate names available under the smtp transport option
1418 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1421 JH/10 Time-related extraction expansions from certificates now use the main
1422 option "timezone" setting for output formatting, and are consistent
1423 between OpenSSL and GnuTLS compilations. Bug 1541.
1425 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1426 encoded parameter in the incoming message. Bug 1558.
1428 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1429 include certificate info, eximon was claiming there were spoolfile
1432 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1434 JH/14 Log delivery-related information more consistently, using the sequence
1435 "H=<name> [<ip>]" wherever possible.
1437 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1438 are problematic for Debian distribution, omit them from the release
1441 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1443 JH/16 Fix string representation of time values on 64bit time_t architectures.
1446 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1447 output list separator was used.
1452 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1453 checkers that were complaining about end of non-void function with no
1456 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1457 This was a regression introduced in 4.83 by another bugfix.
1459 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1461 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1462 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1468 TF/01 Correctly close the server side of TLS when forking for delivery.
1470 When a message was received over SMTP with TLS, Exim failed to clear up
1471 the incoming connection properly after forking off the child process to
1472 deliver the message. In some situations the subsequent outgoing
1473 delivery connection happened to have the same fd number as the incoming
1474 connection previously had. Exim would try to use TLS and fail, logging
1475 a "Bad file descriptor" error.
1477 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1478 utilities have not been installed.
1480 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1481 temporary space as the ACL may create new global variables.
1483 TL/01 LDAP support uses per connection or global context settings, depending
1484 upon the detected version of the libraries at build time.
1486 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1487 to extract and use the src ip:port in logging and expansions as if it
1488 were a direct connection from the outside internet. PPv2 support was
1489 updated based on HAProxy spec change in May 2014.
1491 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1493 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1494 Properly escape header and check for NULL return.
1496 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1499 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1501 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1502 characters in header names, implemented as a verify condition.
1503 Contributed by Michael Fischer v. Mollard.
1505 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1506 results permerror and temperror. Previous values are deprecated but
1507 still accepted. In a future release, err_perm and err_temp will be
1508 completely removed, which will be a backward incompatibility if the
1509 ACL tests for either of these two old results. Patch contributed by
1510 user bes-internal on the mailing list.
1512 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1514 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1515 selectors, in both main and reject logs.
1517 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1520 JH/07 Add malware type "sock" for talking to simple daemon.
1522 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1524 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1525 routers/transports under cutthrough routing.
1527 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1528 numbers. Touch up "bool" conditional to keep the same definition.
1530 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1532 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1534 JH/12 Expand items in router/transport headers_add or headers_remove lists
1535 individually rather than the list as a whole. Bug 1452.
1537 Required for reasonable handling of multiple headers_ options when
1538 they may be empty; requires that headers_remove items with embedded
1539 colons must have them doubled (or the list-separator changed).
1541 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1542 view the policy declared in the DMARC record. Currently, $dmarc_status
1543 is a combined value of both the record presence and the result of the
1546 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1548 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1549 dnslookup router and the smtp transport (applying to the forward
1552 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1553 of ldap servers used for a specific lookup. Patch provided by Heiko
1556 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1557 New variable $lookup_dnssec_authenticated for observability.
1559 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1560 Patch submitted by Lars Timman.
1562 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1564 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1565 Requires trusted mode and valid format message id, aborts otherwise.
1566 Patch contributed by Heiko Schlichting.
1568 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1569 certextract with support for various fields. Bug 1358.
1571 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1572 is requested by default, modifiable by smtp transport option
1575 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1576 operate on certificate variables to give certificate fingerprints
1577 Also new ${sha256:cert_variable}.
1579 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1581 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1582 Christian Aistleitner.
1584 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1586 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1587 file. Patch from Wolfgang Breyha.
1589 JH/25 Expand the coverage of the delivery $host and $host_address to
1590 client authenticators run in verify callout. Bug 1476.
1592 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1593 align with daemon_smtp_ports. Bug 72.
1595 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1596 support and error reporting did not work properly.
1598 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1599 and is readable. Patch from Andrew Colin Kissa.
1601 TL/14 Enhance documentation of ${run expansion and how it parses the
1602 commandline after expansion, particularly in the case when an
1603 unquoted variable expansion results in an empty value.
1605 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1607 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1608 by renaming away from it. Addresses GH issue 15, reported by
1611 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1613 TL/15 SECURITY: prevent double expansion in math comparison functions
1614 (can expand unsanitized data). Not remotely exploitable.
1621 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1623 PP/02 Make -n do something, by making it not do something.
1624 When combined with -bP, the name of an option is not output.
1626 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1629 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1630 $sender_host_name and config options to manage this, and basic check
1633 PP/05 DSCP support for outbound connections and control modifier for inbound.
1635 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1636 (Only plugin which currently uses this is kerberos4, which nobody should
1637 be using, but we should make it available and other future plugins might
1638 conceivably use it, even though it would break NAT; stuff *should* be
1639 using channel bindings instead).
1641 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1642 name; added for Sendmail compatibility; requires admin caller.
1643 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1644 just ignore it); requires trusted caller.
1645 Also parse but ignore: -Ac -Am -X<logfile>
1648 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1650 TL/02 Add +smtp_confirmation as a default logging option.
1652 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1653 Patch by Magnus Holmgren from 2007-02-20.
1655 TL/04 Bugzilla 1281 - Spec typo.
1656 Bugzilla 1283 - Spec typo.
1657 Bugzilla 1290 - Spec grammar fixes.
1659 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1661 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1663 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1664 mailing list by by Dmitry Isaikin.
1666 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1668 JH/02 Support "G" suffix to numbers in ${if comparisons.
1670 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1672 NM/01 Bugzilla 1197 - Spec typo
1673 Bugzilla 1196 - Spec examples corrections
1675 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1677 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1678 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1681 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1682 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1684 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1685 "acl {{name}{arg}...}", and optional args on acl condition
1688 JH/05 Permit multiple router/transport headers_add/remove lines.
1690 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1692 JH/07 Avoid using a waiting database for a single-message-only transport.
1693 Performance patch from Paul Fisher. Bugzilla 1262.
1695 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1698 JH/09 Add $headers_added variable, with content from use of ACL modifier
1699 add_header (but not yet added to the message). Bugzilla 199.
1701 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1702 Pulled from Bugzilla 817 by Wolfgang Breyha.
1704 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1706 (nb: this is the same fix as in Exim 4.80.1)
1708 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1711 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1714 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1716 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1717 advertise SMTP AUTH mechanism to us, instead of a generic
1718 protocol violation error. Also, make Exim more robust to bad
1719 data from the Dovecot auth socket.
1721 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1723 When a queue runner is handling a message, Exim first routes the
1724 recipient addresses, during which it prunes them based on the retry
1725 hints database. After that it attempts to deliver the message to
1726 any remaining recipients. It then updates the hints database using
1729 So if a recipient address works intermittently, it can get repeatedly
1730 deferred at routing time. The retry hints record remains fresh so the
1731 address never reaches the final cutoff time.
1733 This is a fairly common occurrence when a user is bumping up against
1734 their storage quota. Exim had some logic in its local delivery code
1735 to deal with this. However it did not apply to per-recipient defers
1736 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1738 This change adds a proper retry rule check during routing so that the
1739 final cutoff time is checked against the message's age. We only do
1740 this check if there is an address retry record and there is not a
1741 domain retry record; this implies that previous attempts to handle
1742 the address had the retry_use_local_parts option turned on. We use
1743 this as an approximation for the destination being like a local
1744 delivery, as in LMTP.
1746 I suspect this new check makes the old local delivery cutoff check
1747 redundant, but I have not verified this so I left the code in place.
1749 TF/02 Correct gecos expansion when From: is a prefix of the username.
1751 Test 0254 submits a message to Exim with the header
1755 When I ran the test suite under the user fanf2, Exim expanded
1756 the header to contain my full name, whereas it should have added
1757 a Resent-Sender: header. It erroneously treats any prefix of the
1758 username as equal to the username.
1760 This change corrects that bug.
1762 GF/01 DCC debug and logging tidyup
1763 Error conditions log to paniclog rather than rejectlog.
1764 Debug lines prefixed by "DCC: " to remove any ambiguity.
1766 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1768 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1769 Bug spotted by Jeremy Harris; was flawed since initial commit.
1770 Would have resulted in OCSP responses post-SNI triggering an Exim
1771 NULL dereference and crash.
1773 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1775 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1776 Bug detection, analysis and fix by Samuel Thibault.
1777 Bugzilla 1331, Debian bug #698092.
1779 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1781 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1782 Server implementation by Todd Lyons, client by JH.
1783 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1784 config variable "prdr_enable" controls whether the server
1785 advertises the facility. If the client requests PRDR a new
1786 acl_data_smtp_prdr ACL is called once for each recipient, after
1787 the body content is received and before the acl_smtp_data ACL.
1788 The client is controlled by both of: a hosts_try_prdr option
1789 on the smtp transport, and the server advertisement.
1790 Default client logging of deliveries and rejections involving
1791 PRDR are flagged with the string "PRDR".
1793 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1794 fclose(). Diagnosis by Todd Lyons.
1796 PP/17 Update configure.default to handle IPv6 localhost better.
1797 Patch by Alain Williams (plus minor tweaks).
1800 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1801 This is now consistent with GnuTLS, and is now documented: the
1802 previous undocumented portable approach to treating the option as
1803 unset was to force an expansion failure. That still works, and
1804 an empty string is now equivalent.
1806 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1807 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1808 not performing validation itself.
1810 PP/20 Added force_command boolean option to pipe transport.
1811 Patch from Nick Koston, of cPanel Inc.
1813 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1816 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1818 PP/21 Fix eximon continuous updating with timestamped log-files.
1819 Broken in a format-string cleanup in 4.80, missed when I repaired the
1820 other false fix of the same issue.
1821 Report and fix from Heiko Schlichting.
1824 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1825 Report from Prashanth Katuri.
1827 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1828 It's SecureTransport, so affects any MacOS clients which use the
1829 system-integrated TLS libraries, including email clients.
1831 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1832 using a MIME ACL for non-SMTP local injection.
1833 Report and assistance in diagnosis by Warren Baker.
1835 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1837 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1839 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1840 last id that failed so it may be referenced in subsequent ACL's.
1842 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1845 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1846 ldap library initialization, allowing self-signed CA's to be
1847 used. Also properly sets require_cert option later in code by
1848 using NULL (global ldap config) instead of ldap handle (per
1849 session). Bug diagnosis and testing by alxgomz.
1851 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1852 the src/util/ subdirectory.
1854 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1855 renamed to Transport Post Delivery Action by Jeremy Harris, as
1858 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1859 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1860 redis_servers = needs to be configured which will be used by the redis
1861 lookup. Patch from Warren Baker, of The Packet Hub.
1863 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1865 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1866 hostname or reverse DNS when processing a host list. Used suggestions
1867 from multiple comments on this bug.
1869 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1871 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1872 Missed a few lines, added it to make the runtest require no keyboard
1875 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1876 contains upper case chars. Make router use caseful_local_part.
1878 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1879 support when GnuTLS has been built with p11-kit.
1885 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1887 This, or similar/improved, will also be change PP/11 of 4.82.
1893 PP/01 Handle short writes when writing local log-files.
1894 In practice, only affects FreeBSD (8 onwards).
1895 Bugzilla 1053, with thanks to Dmitry Isaikin.
1897 NM/01 Bugzilla 949 - Documentation tweak
1899 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1902 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1904 PP/02 Implemented gsasl authenticator.
1906 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1908 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1909 `pkg-config foo` for cflags/libs.
1911 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1912 with rest of GSASL and with heimdal_gssapi.
1914 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1915 `pkg-config foo` for cflags/libs for the TLS implementation.
1917 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1918 properties get this fed in as external SSF. A number of robustness
1919 and debugging improvements to the cyrus_sasl authenticator.
1921 PP/08 cyrus_sasl server now expands the server_realm option.
1923 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1924 Patch by Jeremy Harris.
1926 PP/10 Added dbmjz lookup type.
1928 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1930 PP/12 MAIL args handles TAB as well as SP, for better interop with
1931 non-compliant senders.
1932 Analysis and variant patch by Todd Lyons.
1934 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1935 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1936 Patch from Dirk Mueller <dmueller@suse.com>
1938 PP/13 tls_peerdn now print-escaped for spool files.
1939 Observed some $tls_peerdn in wild which contained \n, which resulted
1940 in spool file corruption.
1942 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1943 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1944 or write after TLS renegotiation, which otherwise led to messages
1947 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1948 as a tracking header (ie: a signed header comes before the signature).
1949 Patch from Wolfgang Breyha.
1951 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1952 comma-sep list; embedded commas doubled.
1954 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1956 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1958 Report and patch from Dmitry Banschikov.
1960 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1961 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1962 failures. We appear to now support TLS1.1+ with Exim.
1964 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1965 lets Exim select keys and certificates based upon TLS SNI from client.
1966 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1967 before an outbound SMTP session. New log_selector, +tls_sni.
1969 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1970 NULL dereference. Report and patch from Alun Jones.
1972 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1973 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1974 Not seeing resolver debug output on NetBSD, but suspect this is a
1975 resolver implementation change.
1977 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1978 Left warnings. Added "eximon gdb" invocation mode.
1980 PP/21 Defaulting "accept_8bitmime" to true, not false.
1982 PP/22 Added -bw for inetd wait mode support.
1984 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1985 locate the relevant includes and libraries. Made this the default.
1987 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1988 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1990 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1991 This may cause build issues on older platforms.
1993 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1994 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1995 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1996 Added SNI support via GnuTLS too.
1997 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1999 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2001 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2002 Applied second patch from Janne, implementing suggestion to default
2003 multiple-strings-in-record handling to match SPF spec.
2005 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2007 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2008 read-only, out of scope).
2009 Patch from Wolfgang Breyha, report from Stuart Northfield.
2011 PP/29 Fix three issues highlighted by clang analyser static analysis.
2012 Only crash-plausible issue would require the Cambridge-specific
2013 iplookup router and a misconfiguration.
2014 Report from Marcin Mirosław.
2016 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2018 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2019 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2020 As part of this, removing so much warning spew let me fix some minor
2021 real issues in debug logging.
2023 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2024 assignment on my part. Fixed.
2026 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2027 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2028 Janne Snabb (who went above and beyond: thank you).
2030 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2031 string otherwise requires a connection and a bunch more work and it's
2032 relatively easy to get wrong. Should also expose TLS library linkage
2035 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2036 64-bit ${eval} (JH/03).
2038 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2039 GNU libc to support some of the 64-bit stuff, should not lead to
2040 conflicts. Defined before os.h is pulled in, so if a given platform
2041 needs to override this, it can.
2043 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2044 protection layer was required, which is not implemented.
2045 Bugzilla 1254, patch from Wolfgang Breyha.
2047 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2048 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2049 tls_dhparam take prime identifiers. Also unbreak combination of
2050 OpenSSL+DH_params+TLSSNI.
2052 PP/39 Disable SSLv2 by default in OpenSSL support.
2058 PP/01 Solaris build fix for Oracle's LDAP libraries.
2059 Bugzilla 1109, patch from Stephen Usher.
2061 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2063 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2066 TF/02 Fix a couple more cases where we did not log the error message
2067 when unlink() failed. See also change 4.74-TF/03.
2069 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2070 lock up or crash if it happened to be inside a call to libc when it
2071 got a SIGUSR1 from exiwhat.
2073 The SIGUSR1 handler appends the current process status to the process
2074 log which is later printed by exiwhat. It used to use the general
2075 purpose logging code to do this, but several functions it calls are
2076 not safe for signals.
2078 The new output code in the SIGUSR1 handler is specific to the process
2079 log, and simple enough that it's easy to inspect for signal safety.
2080 Removing some special cases also simplifies the general logging code.
2081 Removing the spurious timestamps from the process log simplifies
2084 TF/04 Improved ratelimit ACL condition.
2086 The /noupdate option has been deprecated in favour of /readonly which
2087 has clearer semantics. The /leaky, /strict, and /readonly update modes
2088 are mutually exclusive. The update mode is no longer included in the
2089 database key; it just determines when the database is updated. (This
2090 means that when you upgrade Exim will forget old rate measurements.)
2092 Exim now checks that the per_* options are used with an update mode that
2093 makes sense for the current ACL. For example, when Exim is processing a
2094 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2095 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2096 must specify per_mail/readonly. If you omit the update mode it defaults to
2097 /leaky where that makes sense (as before) or /readonly where required.
2099 The /noupdate option is now undocumented but still supported for
2100 backwards compatibility. It is equivalent to /readonly except that in
2101 ACLs where /readonly is required you may specify /leaky/noupdate or
2102 /strict/noupdate which are treated the same as /readonly.
2104 A useful new feature is the /count= option. This is a generalization
2105 of the per_byte option, so that you can measure the throughput of other
2106 aggregate values. For example, the per_byte option is now equivalent
2107 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2109 The per_rcpt option has been generalized using the /count= mechanism
2110 (though it's more complicated than the per_byte equivalence). When it is
2111 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2112 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2113 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2114 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2115 Note that using per_rcpt with a non-readonly update mode in more than
2116 one ACL will cause the recipients to be double-counted. (The per_mail
2117 and per_byte options don't have this problem.)
2119 The handling of very low rates has changed slightly. If the computed rate
2120 is less than the event's count (usually one) then this event is the first
2121 after a long gap. In this case the rate is set to the same as this event's
2122 count, so that the first message of a spam run is counted properly.
2124 The major new feature is a mechanism for counting the rate of unique
2125 events. The new per_addr option counts the number of different
2126 recipients that someone has sent messages to in the last time period. It
2127 behaves like per_rcpt if all the recipient addresses are different, but
2128 duplicate recipient addresses do not increase the measured rate. Like
2129 the /count= option this is a general mechanism, so the per_addr option
2130 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2131 example, measure the rate that a client uses different sender addresses
2132 with the options per_mail/unique=$sender_address. There are further
2133 details in the main documentation.
2135 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2137 TF/06 Removed a few PCRE remnants.
2139 TF/07 Automatically extract Exim's version number from tags in the git
2140 repository when doing development or release builds.
2142 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2143 Bugzilla 879. Patch from Paul Fisher.
2145 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2146 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2149 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2151 PP/05 Variable $av_failed, true if the AV scanner deferred.
2152 Bugzilla 1078. Patch from John Horne.
2154 PP/06 Stop make process more reliably on build failure.
2155 Bugzilla 1087. Patch from Heiko Schlittermann.
2157 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2158 Bugzilla 1089. Patch from Heiko Schlittermann.
2160 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2161 Bugzilla 1131. Patch from Holger Weiß.
2163 PP/09 Handle IPv6 addresses with SPF.
2164 Bugzilla 860. Patch from Wolfgang Breyha.
2166 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2168 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2171 PP/11 match_* no longer expand right-hand-side by default.
2172 New compile-time build option, EXPAND_LISTMATCH_RHS.
2173 New expansion conditions, "inlist", "inlisti".
2175 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2177 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2179 PP/14 fix log_write() format string regression from TF/03.
2180 Bugzilla 1152. Patch from Dmitry Isaikin.
2186 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2188 PP/02 Harmonised TLS library version reporting; only show if debugging.
2189 Layout now matches that introduced for other libraries in 4.74 PP/03.
2191 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2193 PP/04 New "dns_use_edns0" global option.
2195 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2198 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2199 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2201 TK/01 Updated PolarSSL code to 0.14.2.
2202 Bugzilla 1097. Patch from Andreas Metzler.
2204 PP/07 Catch divide-by-zero in ${eval:...}.
2205 Fixes bugzilla 1102.
2207 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2210 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2211 format-string attack -- SECURITY: remote arbitrary code execution.
2213 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2214 time unintentionally subject to list matching rules, letting the header
2215 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2216 arbitrary string expansion). This allowed for information disclosure.
2218 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2219 INT_MIN/-1 -- value coerced to INT_MAX.
2225 NM/01 Workaround for PCRE version dependency in version reporting
2228 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2229 This fixes portability to compilers other than gcc, notably
2230 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2232 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2233 makefiles for portability to HP-UX and POSIX correctness.
2235 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2236 Also via indented variable definition in the Makefile.
2237 (Debugging by Oliver Heesakkers).
2239 PP/02 Restore caching of spamd results with expanded spamd_address.
2240 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2242 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2243 Improves build reliability. Fix from: Frank Elsner
2245 NM/02 Fix wide character breakage in the rfc2047 coding
2246 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2248 NM/03 Allow underscore in dnslist lookups
2249 Fixes bug 1026. Patch from Graeme Fowler
2251 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2252 Code patches from Adam Ciarcinski of NetBSD.
2254 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2257 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2258 is logged, to avoid truncation. Patch from John Horne.
2260 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2261 Patch from Jakob Hirsch.
2263 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2264 SQL string expansion failure details.
2265 Patch from Andrey Oktyabrski.
2267 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2268 Patch from Simon Arlott.
2270 PP/09 New lookups functionality failed to compile on old gcc which rejects
2271 extern declarations in function scope.
2272 Patch from Oliver Fleischmann
2274 PP/10 Use sig_atomic_t for flags set from signal handlers.
2275 Check getgroups() return and improve debugging.
2276 Fixed developed for diagnosis in bug 927 (which turned out to be
2279 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2280 Patch from Mark Zealey.
2282 PP/12 Bugzilla 1056: Improved spamd server selection.
2283 Patch from Mark Zealey.
2285 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2286 Based on patch from Heiko Schlittermann.
2288 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2289 Patch from Uwe Doering, sign-off by Michael Haardt.
2291 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2292 variable declaration deep within a block. Bug and patch from
2295 PP/15 lookups-Makefile IRIX compatibility coercion.
2297 PP/16 Make DISABLE_DKIM build knob functional.
2299 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2300 Patch by Simon Arlott
2302 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2303 variable argument macros. Our copy now differs from upstream.
2309 TF/01 Failure to get a lock on a hints database can have serious
2310 consequences so log it to the panic log.
2312 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2313 controlled using the smtp_confirmation log selector.
2315 TF/03 Include the error message when we fail to unlink a spool file.
2317 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2318 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2319 for maintaining out-of-tree patches for some time.
2321 PP/01 Bugzilla 139: Documentation and portability issues.
2322 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2323 Handle per-OS dynamic-module compilation flags.
2325 PP/02 Let /dev/null have normal permissions.
2326 The 4.73 fixes were a little too stringent and complained about the
2327 permissions on /dev/null. Exempt it from some checks.
2328 Reported by Andreas M. Kirchwitz.
2330 PP/03 Report version information for many libraries, including
2331 Exim version information for dynamically loaded libraries. Created
2332 version.h, now support a version extension string for distributors
2333 who patch heavily. Dynamic module ABI change.
2335 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2336 privilege escalation vulnerability whereby the Exim run-time user
2337 can cause root to append content of the attacker's choosing to
2340 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2343 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2344 If dropping privileges for untrusted macros, we disabled normal logging
2345 on the basis that it would fail; for the Exim run-time user, this is not
2346 the case, and it resulted in successful deliveries going unlogged.
2347 Fixed. Reported by Andreas Metzler.
2353 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2354 only prepend for the Resent-* case. Fixes regression introduced in
2355 Exim 4.70 by NM/22 for Bugzilla 607.
2357 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2358 increasing numbers of administrators be bitten by this.
2360 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2362 PP/03 Bugzilla 994: added openssl_options main configuration option.
2364 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2366 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2368 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2370 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2371 without a peer certificate, leading to a segfault because of an
2372 assumption that peers always have certificates. Be a little more
2373 paranoid. Problem reported by Martin Tscholak.
2375 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2376 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2377 NB: ClamAV planning to remove STREAM in "middle of 2010".
2378 CL also introduces -bmalware, various -d+acl logging additions and
2379 more caution in buffer sizes.
2381 PP/09 Implemented reverse_ip expansion operator.
2383 PP/10 Bugzilla 937: provide a "debug" ACL control.
2385 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2387 PP/12 Bugzilla 973: Implement --version.
2389 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2391 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2393 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2395 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2396 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2397 ignore trailing whitespace.
2399 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2401 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2404 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2405 Notification from Dr Andrew Aitchison.
2407 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2408 ExtendedDetectionInfo response format.
2409 Notification from John Horne.
2411 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2414 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2415 XSL and documented dependency on system catalogs, with examples of how
2418 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2421 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2422 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2423 configuration file which is writeable by the Exim user or group.
2425 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2426 of configuration files to cover files specified with the -C option if
2427 they are going to be used with root privileges, not just the default
2430 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2431 option (effectively making it always true).
2433 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2434 files to be used while preserving root privileges.
2436 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2437 that rogue child processes cannot use them.
2439 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2440 run-time user, instead of root.
2442 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2443 Exim run-time user without dropping privileges.
2445 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2446 result string, instead of calling string_vformat() twice with the same
2449 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2450 for other users. Others should always drop root privileges if they use
2451 -C on the command line, even for a whitelisted configure file.
2453 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2455 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2461 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2462 $data_path, and $header_path variables; fixed documentation bugs and
2465 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2466 exipick to access non-standard spools, including the "frozen" queue
2469 NM/01 Bugzilla 965: Support mysql stored procedures.
2470 Patch from Alain Williams
2472 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2474 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2475 Patch from Andreas Metzler
2477 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2478 Patch from Kirill Miazine
2480 NM/05 Bugzilla 671: Added umask to procmail example.
2482 JJ/03 installed exipick 20100323.0, fixing doc bug
2484 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2485 directory. Notification and patch from Dan Rosenberg.
2487 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2489 TK/02 Improve log output when DKIM signing operation fails.
2491 MH/01 Treat the transport option dkim_domain as a colon separated
2492 list, not as a single string, and sign the message with each element,
2493 omitting multiple occurences of the same signer.
2495 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2496 Bugzilla 985, 986. Patch by Simon Arlott
2498 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2499 Patch by Simon Arlott
2501 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2502 MBX locking. Notification from Dan Rosenberg.
2508 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2510 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2512 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2514 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2516 NM/04 Bugzilla 915: Flag broken perl installation during build.
2522 TK/01 Added patch by Johannes Berg that expands the main option
2523 "spamd_address" if it starts with a dollar sign.
2525 TK/02 Write list of recipients to X-Envelope-Sender header when building
2526 the mbox-format spool file for content scanning (suggested by Jakob
2529 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2530 (http://www.dcc-servers.net/) support via dccifd. Activated by
2531 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2533 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2534 by Mark Daniel Reidel <mr@df.eu>.
2536 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2537 When building exim an external PCRE library is now needed -
2538 PCRE is a system library on the majority of modern systems.
2539 See entry on PCRE_LIBS in EDITME file.
2541 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2542 conversation. Added nologin parameter to request.
2543 Patch contributed by Kirill Miazine.
2545 TF/01 Do not log submission mode rewrites if they do not change the address.
2547 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2549 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2550 log files in place. Contributed by Roberto Lima.
2552 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2554 TF/03 Bugzilla 615: When checking the local_parts router precondition
2555 after a local_part_suffix or local_part_prefix option, Exim now
2556 does not use the address's named list lookup cache, since this
2557 contains cached lookups for the whole local part.
2559 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2560 Robert Millan. Documentation is in experimental-spec.txt.
2562 TF/04 Bugzilla 668: Fix parallel build (make -j).
2564 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2566 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2567 Patch provided by Jan Srzednicki.
2569 TF/05 Leading white space used to be stripped from $spam_report which
2570 wrecked the formatting. Now it is preserved.
2572 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2573 that they are available at delivery time.
2575 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2577 TF/08 TLS error reporting now respects the incoming_interface and
2578 incoming_port log selectors.
2580 TF/09 Produce a more useful error message if an SMTP transport's hosts
2581 setting expands to an empty string.
2583 NM/06 Bugzilla 744: EXPN did not work under TLS.
2584 Patch provided by Phil Pennock.
2586 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2587 Patch provided by Richard Godbee.
2589 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2590 acl_smtp_notquit, added index entry.
2592 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2593 Patch provided by Eugene Bujak.
2595 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2596 accept(). Patch provided by Maxim Dounin.
2598 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2599 Patch provided by Phil Pennock.
2601 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2603 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2604 Patch provided by Brad "anomie" Jorsch.
2606 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2607 Patch provided by Dean Brooks.
2609 TK/05 Add native DKIM support (does not depend on external libraries).
2611 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2612 Patch provided by Graeme Fowler.
2614 NM/16 Bugzilla 851: Documentation example syntax fix.
2616 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2618 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2621 NM/19 Bugzilla 745: TLS version reporting.
2622 Patch provided by Phil Pennock.
2624 NM/20 Bugzilla 167: bool: condition support.
2625 Patch provided by Phil Pennock.
2627 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2628 clients. Patch provided by Phil Pennock.
2630 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2631 Patch provided by Brad "anomie" Jorsch.
2633 NM/23 Bugzilla 687: Fix misparses in eximstats.
2634 Patch provided by Heiko Schlittermann.
2636 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2637 Patch provided by Heiko Schlittermann.
2639 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2640 plus update to original patch.
2642 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2644 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2645 Patch provided by David Brownlee.
2647 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2649 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2651 NM/30 Bugzilla 888: TLS documentation bugfixes.
2653 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2655 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2656 Unlike the original bugzilla I have changed all shell scripts in src tree.
2658 NM/33 Bugzilla 898: Transport filter timeout fix.
2659 Patch by Todd Rinaldo.
2661 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2662 Patch by Serge Demonchaux.
2664 NM/35 Bugzilla 39: Base64 decode bug fixes.
2665 Patch by Jakob Hirsch.
2667 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2669 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2671 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2673 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2679 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2680 ALT-N's libdkim that I have put here:
2681 http://duncanthrax.net/exim-experimental/
2683 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2684 were called 'true' and it seems that C99 defines that as a reserved
2685 keyword to be used with 'bool' variable types. That means you could
2686 not include C99-style headers which use bools without triggering
2687 build errors in sieve.c.
2689 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2690 as mailq or other aliases. Changed the --help handling significantly
2691 to do whats expected. exim_usage() emits usage/help information.
2693 SC/01 Added the -bylocaldomain option to eximstats.
2695 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2697 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2699 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2705 PH/01 Another patch from the Sieve maintainer.
2707 PH/02 When an IPv6 address is converted to a string for single-key lookup
2708 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2709 dots are used instead of colons so that keys in lsearch files need not
2710 contain colons. This was done some time before quoting was made available
2711 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2712 (notated using the quote facility) so as to distinguish them from IPv4
2713 keys. This meant that lookups for IP addresses in host lists did not work
2714 for iplsearch lookups.
2716 This has been fixed by arranging for IPv6 addresses to be expressed with
2717 colons if the lookup type is iplsearch. This is not incompatible, because
2718 previously such lookups could never work.
2720 The situation is now rather anomalous, since one *can* have colons in
2721 ordinary lsearch keys. However, making the change in all cases is
2722 incompatible and would probably break a number of configurations.
2724 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2727 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2728 conversion specification without a maximum field width, thereby enabling
2729 a rogue spamd server to cause a buffer overflow. While nobody in their
2730 right mind would setup Exim to query an untrusted spamd server, an
2731 attacker that gains access to a server running spamd could potentially
2732 exploit this vulnerability to run arbitrary code as the Exim user.
2734 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2735 $primary_hostname instead of what libspf2 thinks the hosts name is.
2737 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2738 a directory entry by the name of the lookup key. Previously, if a
2739 symlink pointed to a non-existing file or a file in a directory that
2740 Exim lacked permissions to read, a lookup for a key matching that
2741 symlink would fail. Now it is enough that a matching directory entry
2742 exists, symlink or not. (Bugzilla 503.)
2744 PH/03 The body_linecount and body_zerocount variables are now exported in the
2747 PH/04 Added the $dnslist_matched variable.
2749 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2750 This means they are set thereafter only if the connection becomes
2753 PH/06 Added the client_condition to authenticators so that some can be skipped
2754 by clients under certain conditions.
2756 PH/07 The error message for a badly-placed control=no_multiline_responses left
2757 "_responses" off the end of the name.
2759 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2761 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2762 (without spaces) instead of just copying the configuration text.
2764 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2766 PH/11 Added $max_received_linelength.
2768 PH/12 Added +ignore_defer and +include_defer to host lists.
2770 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2771 way in which PCRE > 7.0 is built.
2773 PH/14 Implemented queue_only_load_latch.
2775 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2776 MAIL command. The effect was to mangle the value on 64-bit systems.
2778 PH/16 Another patch from the Sieve maintainer.
2780 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2782 PH/18 If a system quota error occurred while trying to create the file for
2783 a maildir delivery, the message "Mailbox is full" was not appended to the
2784 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2785 only to a quota excession during the actual writing of the file.
2787 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2788 characters?) which causes problems in log lines. The DN values are now
2789 passed through string_printing() before being added to log lines.
2791 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2792 and InterBase are left for another time.)
2794 PH/21 Added message_body_newlines option.
2796 PH/22 Guard against possible overflow in moan_check_errorcopy().
2798 PH/23 POSIX allows open() to be a macro; guard against that.
2800 PH/24 If the recipient of an error message contained an @ in the local part
2801 (suitably quoted, of course), incorrect values were put in $domain and
2802 $local_part during the evaluation of errors_copy.
2808 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2809 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2812 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2813 issue a MAIL command.
2815 PH/02 In an ACL statement such as
2817 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2819 if a client was not listed at all, or was listed with a value other than
2820 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2821 the condition was not true (as it should be), so access was not denied.
2822 The bug was that the ! inversion was incorrectly passed on to the second
2823 item. This has been fixed.
2825 PH/03 Added additional dnslists conditions == and =& which are different from
2826 = and & when the dns lookup returns more than one IP address.
2828 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2829 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2831 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2832 FSYNC, which compiles an option called disable_fsync that allows for
2833 bypassing fsync(). The documentation is heavily laced with warnings.
2835 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2837 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2838 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2839 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2840 including adding "make clean"; (3) Added -fPIC when compiling the test
2841 dynamically loaded module, to get rid of a warning.
2843 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2844 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2845 = 0s. The bug is otherwise harmless.
2847 PH/07 There was a bug in the dovecot authenticator such that the value of
2848 $auth1 could be overwritten, and so not correctly preserved, after a
2849 successful authentication. This usually meant that the value preserved by
2850 the server_setid option was incorrect.
2852 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2854 PH/09 Installed PCRE release 7.0.
2856 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2857 run for batched SMTP input. It is now run at the start of every message
2858 in the batch. While fixing this I discovered that the process information
2859 (output by running exiwhat) was not always getting set for -bs and -bS
2860 input. This is fixed, and it now also says "batched" for BSMTP.
2862 PH/11 Added control=no_pipelining.
2864 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2865 patch, slightly modified), and move the expansion of helo_data till after
2866 the connection is made in the smtp transport (so it can use these
2869 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2871 PH/14 Added log_selector = +pid.
2873 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2875 PH/16 Add ${if forany and ${if forall.
2877 PH/17 Added dsn_from option to vary the From: line in DSNs.
2879 PH/18 Flush SMTP output before performing a callout, unless control =
2880 no_callout_flush is set.
2882 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2883 was true (the default) a successful delivery failed to delete the retry
2884 item, thus causing premature timeout of the address. The bug is now
2887 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2889 PH/21 Long custom messages for fakedefer and fakereject are now split up
2890 into multiline responses in the same way that messages for "deny" and
2891 other ACL rejections are.
2893 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2894 with slight modification.
2896 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2897 draft, changing the syntax and factoring some duplicate code".
2899 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2900 for deliveries of the second and subsequent messages over the same SMTP
2903 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2904 ${reduce, with only minor "tidies".
2906 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2908 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2909 expansion side effects.
2911 PH/27 When a message times out after an over-quota error from an Exim-imposed
2912 quota, the bounce message says "mailbox is full". This message was not
2913 being given when it was a system quota that was exceeded. It now should
2916 MH/03 Made $recipients available in local_scan(). local_scan() already has
2917 better access to the recipient list through recipients_list[], but
2918 $recipients can be useful in postmaster-provided expansion strings.
2920 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2921 in the case of a MAIL command with additional options following the
2922 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2923 were accidentally chopped off.
2925 PH/29 SMTP synchronization checks are implemented when a command is read -
2926 there is a check that no more input is waiting when there shouldn't be
2927 any. However, for some commands, a delay in an ACL can mean that it is
2928 some time before the response is written. In this time, more input might
2929 arrive, invalidly. So now there are extra checks after an ACL has run for
2930 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2931 pipelining has not been advertised.
2933 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2935 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2936 returned in the SMTP response when smtp_return_error_details was false.
2937 This has been fixed.
2939 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2940 instead of the C I/O that was originally supplied, because problems were
2941 reported on Solaris.
2943 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2944 Exim which did not show up earlier: it was assuming that a call to
2945 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2946 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2947 was a macro that became an assignment, so it seemed to work. This has
2948 changed to a proper function call with a void return, hence the compile
2949 error. Exim's code has been fixed.
2951 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2954 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2956 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2958 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2959 because it turns out to be too big or there is a timeout) there is no
2960 "Completed" line in the log. When some messages of this type were
2961 selected by exigrep, they were listed as "not completed". Others were
2962 picked up by some special patterns. I have improved the selection
2963 criteria to be more general.
2965 PH/38 The host_find_failed option in the manualroute router can now be set
2966 to "ignore", to completely ignore a host whose IP address cannot be
2967 found. If all hosts are ignored, the behaviour is controlled by the new
2968 host_all_ignored option.
2970 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2971 homing or because of multiple MX records with /mx) generated more than
2972 one IP address, and the following item turned out to be the local host,
2973 all the secondary addresses of the first item were incorrectly removed
2974 from the list, along with the local host and any following hosts (which
2975 is what is supposed to happen).
2977 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2978 whoever called Exim into the -H file. In the case of the daemon it was
2979 behaving confusingly. When first started, it used values for whoever
2980 started the daemon, but after a SIGHUP it used the Exim user (because it
2981 calls itself on a restart). I have changed the code so that it now always
2984 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2985 message are rejected with the same error (e.g. no authentication or bad
2986 sender address), and a DATA command is nevertheless sent (as can happen
2987 with PIPELINING or a stupid MUA), the error message that was given to the
2988 RCPT commands is included in the rejection of the DATA command. This is
2989 intended to be helpful for MUAs that show only the final error to their
2992 PH/42 Another patch from the Sieve maintainer.
2994 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2995 Eximstats - Fixed some broken HTML links and added missing column headers
2997 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2998 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3000 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3002 PH/43 Yet another patch from the Sieve maintainer.
3004 PH/44 I found a way to check for a TCP/IP connection going away before sending
3005 the response to the final '.' that terminates a message, but only in the
3006 case where the client has not sent further data following the '.'
3007 (unfortunately, this is allowed). However, in many cases there won't be
3008 any further data because there won't be any more messages to send. A call
3009 to select() can be used: if it shows that the input is "ready", there is
3010 either input waiting, or the socket has been closed. An attempt to read
3011 the next input character can distinguish the two cases. Previously, Exim
3012 would have sent an OK response which the client would never have see.
3013 This could lead to message repetition. This fix should cure that, at
3014 least in a lot of common cases.
3016 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3017 advertised in response to EHLO.
3023 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3024 fixed by 4.65/MH/01 (is this a record?) are fixed:
3026 (i) An empty string was always treated as zero by the numeric comparison
3027 operators. This behaviour has been restored.
3029 (ii) It is documented that the numeric comparison operators always treat
3030 their arguments as decimal numbers. This was broken in that numbers
3031 starting with 0 were being interpreted as octal.
3033 While fixing these problems I realized that there was another issue that
3034 hadn't been noticed. Values of message_size_limit (both the global option
3035 and the transport option) were treated as octal if they started with 0.
3036 The documentation was vague. These values are now always treated as
3037 decimal, and I will make that clear in the documentation.
3043 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3044 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3047 MH/01 Don't check that the operands of numeric comparison operators are
3048 integers when their expansion is in "skipping" mode (fixes bug
3049 introduced by 4.64-PH/07).
3051 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3052 child addresses, Exim now panics and dies. Previously, because the count
3053 is held in a short int, deliveries were likely to be lost. As such a
3054 large number of recipients for a single message is ridiculous
3055 (performance will be very, very poor), I have chosen to impose a limit
3056 rather than extend the field.
3062 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3063 leftover -K file (the existence of which was triggered by #402).
3064 While we were at it, introduced process PID as part of the -K
3065 filename. This should rule out race conditions when creating
3068 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3069 processing considerably. Previous code took too long for large mails,
3070 triggering a timeout which in turn triggers #401.
3072 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3073 in the DK code in transports.c. sendfile() is not really portable,
3074 hence the _LINUX specificness.
3076 TF/01 In the add_headers option to the mail command in an Exim filter,
3077 there was a bug that Exim would claim a syntax error in any
3078 header after the first one which had an odd number of characters
3081 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3082 callout verification, Exim cached a "reject" for the entire domain. This
3083 is correct for most verifications, but it is not correct for a recipient
3084 verification with use_sender or use_postmaster set, because in that case
3085 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3086 case of MAIL FROM:<> rejection from other early rejections (e.g.
3087 rejection of HELO). When verifying a recipient using a non-null MAIL
3088 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3089 Whatever the result of the callout, the value of the domain cache is
3090 left unchanged (for any other kind of callout, getting as far as trying
3091 RCPT means that the domain itself is ok).
3093 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3096 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3097 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3098 session, the daemon ignores ECONNECT errors and logs others; it now
3099 ignores EPIPE as well.
3101 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3102 (quoted-printable decoding).
3104 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3105 later the small subsequent patch to fix an introduced bug.
3107 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3109 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3111 PH/08 An error is now given if message_size_limit is specified negative.
3113 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3114 to be given (somewhat) arbitrary names.
3116 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3119 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3120 miscellaneous code fixes
3122 PH/10 Added the log_reject_target ACL modifier to specify where to log
3125 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3126 hostname. This is wrong, because it relates to the incoming message (and
3127 probably the interface on which it is arriving) and not to the outgoing
3128 callout (which could be using a different interface). This has been
3129 changed to use the value of the helo_data option from the smtp transport
3130 instead - this is what is used when a message is actually being sent. If
3131 there is no remote transport (possible with a router that sets up host
3132 addresses), $smtp_active_hostname is used.
3134 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3135 tweaks were necessary in order to get it to work (see also 21 below):
3136 (a) The code assumed that strncpy() returns a negative number on buffer
3137 overflow, which isn't the case. Replaced with Exim's string_format()
3139 (b) There were several signed/unsigned issues. I just did the minimum
3140 hacking in of casts. There is scope for a larger refactoring.
3141 (c) The code used strcasecmp() which is not a standard C function.
3142 Replaced with Exim's strcmpic() function.
3143 (d) The code set only $1; it now sets $auth1 as well.
3144 (e) A simple test gave the error "authentication client didn't specify
3145 service in request". It would seem that Dovecot has changed its
3146 interface. Fortunately there's a specification; I followed it and
3147 changed what the client sends and it appears to be working now.
3149 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3152 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3153 address A is aliased to B and C, where B exists and C does not. Without
3154 -v the output is "A verified" because verification stops after a
3155 successful redirection if more than one address is generated. However,
3156 with -v the child addresses are also verified. Exim was outputting "A
3157 failed to verify" and then showing the successful verification for C,
3158 with its parentage. It now outputs "B failed to verify", showing B's
3159 parentage before showing the successful verification of C.
3161 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3162 look up a TXT record in a specific list after matching in a combined
3165 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3166 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3167 they consult the DNS. I had assumed they would set it the way they
3168 wanted; and indeed my experiments on Linux seem to show that in some
3169 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3170 To be on the safe side, however, I have now made the interface to
3171 host_find_byname() similar to host_find_bydns(), with an argument
3172 containing the DNS resolver options. The host_find_byname() function now
3173 sets these options at its start, just as host_find_bydns() does. The smtp
3174 transport options dns_qualify_single and dns_search_parents are passed to
3175 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3176 of host_find_byname() use the default settings of RES_DEFNAMES
3177 (qualify_single) but not RES_DNSRCH (search_parents).
3179 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3180 spool_read_header() do less string testing, by means of a preliminary
3181 switch on the second character of optional "-foo" lines. (This is
3182 overdue, caused by the large number of possibilities that now exist.
3183 Originally there were few.) While I was there, I also converted the
3184 str(n)cmp tests so they don't re-test the leading "-" and the first
3185 character, in the hope this might squeeze out yet more improvement.
3187 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3188 flag allowing group syntax was set by the header_syntax check but not
3189 turned off, possible causing trouble later; (2) The flag was not being
3190 set at all for the header_verify test, causing "group"-style headers to
3191 be rejected. I have now set it in this case, and also caused header_
3192 verify to ignore an empty address taken from a group. While doing this, I
3193 came across some other cases where the code for allowing group syntax
3194 while scanning a header line wasn't quite right (mostly, not resetting
3195 the flag correctly in the right place). These bugs could have caused
3196 trouble for malformed header lines. I hope it is now all correct.
3198 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3199 with the "reply" argument non-NULL. The code, however (which originally
3200 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3201 but it didn't always do it. This confused somebody who was copying the
3202 code for some other use. I have removed all the tests.
3204 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3205 feature that was used to support insecure browsers during the U.S. crypto
3206 embargo. It requires special client support, and Exim is probably the
3207 only MTA that supported it -- and would never use it because real RSA is
3208 always available. This code has been removed, because it had the bad
3209 effect of slowing Exim down by computing (never used) parameters for the
3210 RSA_EXPORT functionality.
3212 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3213 authenticator to fail if there's a tab character in the incoming data
3214 (there should never be unless someone is messing about, as it's supposed
3215 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3216 if the connection is using TLS or if the remote IP is the same as the
3217 local IP, and the "valid-client-cert option" if a client certificate has
3220 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3221 authenticators. This can be used for authorization after authentication
3222 succeeds. (In the case of plaintext, it servers for both authentication
3225 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3226 if any retry times were supplied.
3228 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3229 connection, where there is no client IP address to check. In this
3230 situation, the verify now always succeeds.
3232 PH/25 Applied John Jetmore's -Mset patch.
3234 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3236 PH/27 In a string expansion for a processed (not raw) header when multiple
3237 headers of the same name were present, leading whitespace was being
3238 removed from all of them, but trailing whitespace was being removed only
3239 from the last one. Now trailing whitespace is removed from each header
3240 before concatenation. Completely empty headers in a concatenation (as
3241 before) are ignored.
3243 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3244 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3246 PH/29 [Removed. This was a change that I later backed out, and forgot to
3247 correct the ChangeLog entry (that I had efficiently created) before
3248 committing the later change.]
3250 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3251 address errors (4xx response to RCPT) over the same connection as other
3252 messages routed to the same hosts. Such deliveries are always "forced",
3253 so retry times are not inspected. This resulted in far too many retries
3254 for the affected addresses. The effect occurred only when there were more
3255 hosts than the hosts_max_try setting in the smtp transport when it had
3256 the 4xx errors. Those hosts that it had tried were not added to the list
3257 of hosts for which the message was waiting, so if all were tried, there
3258 was no problem. Two fixes have been applied:
3260 (i) If there are any address or message errors in an SMTP delivery, none
3261 of the hosts (tried or untried) are now added to the list of hosts
3262 for which the message is waiting, so the message should not be a
3263 candidate for sending over the same connection that was used for a
3264 successful delivery of some other message. This seems entirely
3265 reasonable: after all the message is NOT "waiting for some host".
3266 This is so "obvious" that I'm not sure why it wasn't done
3267 previously. Hope I haven't missed anything, but it can't do any
3268 harm, as the worst effect is to miss an optimization.
3270 (ii) If, despite (i), such a delivery is accidentally attempted, the
3271 routing retry time is respected, so at least it doesn't keep
3272 hammering the server.
3274 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3275 in ${readsocket because some servers need this prod.
3277 PH/32 Added some extra debug output when updating a wait-xxx database.
3279 PH/33 The hint "could be header name not terminated by colon", which has been
3280 given for certain expansion errors for a long time, was not being given
3281 for the ${if def:h_colon_omitted{... case.
3283 PH/34 The spec says: "With one important exception, whenever a domain list is
3284 being scanned, $domain contains the subject domain." There was at least
3285 one case where this was not true.
3287 PH/35 The error "getsockname() failed: connection reset by peer" was being
3288 written to the panic log as well as the main log, but it isn't really
3289 panic-worthy as it just means the connection died rather early on. I have
3290 removed the panic log writing for the ECONNRESET error when getsockname()
3293 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3294 runs only) independently of the message's sender address. This meant
3295 that, if the 4xx error was in fact related to the sender, a different
3296 message to the same recipient with a different sender could confuse
3297 things. In particular, this can happen when sending to a greylisting
3298 server, but other circumstances could also provoke similar problems.
3299 I have changed the default so that the retry time for these errors is now
3300 based a combination of the sender and recipient addresses. This change
3301 can be overridden by setting address_retry_include_sender=false in the
3304 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3305 remote server are returned as part of bounce messages. This was not
3306 happening for LMTP over a pipe (the lmtp transport), but now it is the
3307 same for both kinds of LMTP.
3309 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3310 in header lines that were in fact CNAMEs. This is no longer the case.
3312 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3313 and queue runs started by the daemon processed all messages. This has
3314 been fixed so that -R and -S can now usefully be given with -q<time>.
3316 PH/40 Import PCRE release 6.7 (fixes some bugs).
3318 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3320 PH/42 Give an error if -q is specified more than once.
3322 PH/43 Renamed the variables $interface_address and $interface_port as
3323 $received_ip_address and $received_port, to make it clear that these
3324 values apply to message reception, and not to the outgoing interface when
3325 a message is delivered. (The old names remain recognized, of course.)
3327 PH/44 There was no timeout on the connect() call when using a Unix domain
3328 socket in the ${readsocket expansion. There now is.
3330 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3331 be meaningful with "accept".
3333 SC/01 Eximstats V1.43
3334 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3336 SC/02 Eximstats V1.44
3337 Use a glob alias rather than an array ref in the generated
3338 parser. This improves both readability and performance.
3340 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3341 Collect SpamAssassin and rejection statistics.
3342 Don't display local sender or destination tables unless
3343 there is data to show.
3344 Added average volumes into the top table text output.
3346 SC/04 Eximstats V1.46
3347 Collect data on the number of addresses (recipients)
3348 as well as the number of messages.
3350 SC/05 Eximstats V1.47
3351 Added 'Message too big' to the list of mail rejection
3352 reasons (thanks to Marco Gaiarin).
3354 SC/06 Eximstats V1.48
3355 Mainlog lines which have GMT offsets and are too short to
3356 have a flag are now skipped.
3358 SC/07 Eximstats V1.49 (Alain Williams)
3359 Added the -emptyok flag.
3361 SC/08 Eximstats V1.50
3362 Fixes for obtaining the IP address from reject messages.
3364 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3365 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3366 whitespace changes from 4.64-PH/27
3368 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3371 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3372 are found, allow negative numbers in numeric criteria)
3374 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3376 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3377 to match changes made in 4.64-PH/43
3379 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3381 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3382 "domain" is actually an IP address, and force a failure. This locks out
3383 those revolvers/nameservers that support "A-for-A" lookups, in
3384 contravention of the specifications.
3386 PH/48 When a host name was looked up from an IP address, and the subsequent
3387 forward lookup of the name timed out, the host name was left in
3388 $sender_host_name, contrary to the specification.
3390 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3391 restricted to single-key lookups, Exim was not diagnosing an error if
3392 * or *@ was used with a query-style lookup.
3394 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3396 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3397 long ago, but noone interested enough thought of it. Let's just say that
3398 the "1.1" means that there are some new functions that weren't there at
3399 some point in the past.
3401 PH/51 Error processing for expansion failure of helo_data from an smtp
3402 transport during callout processing was broken.
3404 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3405 tested/used via the -bh/-bhc/-bs options.
3407 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3408 bug, fixed in subsequent PCRE releases).
3410 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3411 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3413 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3419 SC/01 Use a glob alias rather than an array ref in eximstats generated
3420 parser. This improves both readability and performance.
3422 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3423 Don't display local sender or destination tables in eximstats unless
3424 there is data to show.
3425 Added average volumes into the eximstats top table text output.
3427 SC/03 Collect data on the number of addresses (recipients) as well
3428 as the number of messages in eximstats.
3430 TF/01 Correct an error in the documentation for the redirect router. Exim
3431 does (usually) call initgroups() when daemonizing.
3433 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3434 with consistent privilege compared to when running as a daemon.
3436 TF/03 Note in the spec that $authenticated_id is not set for local
3437 submissions from trusted users.
3439 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3440 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3442 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3443 by adding some example configuration directives to the default
3444 configuration file. A little bit of work is required to uncomment the
3445 directives and define how usernames and passwords are checked, but
3446 there is now a framework to start from.
3448 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3449 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3450 without this. I don't know how relevant this is to other LDAP libraries.
3452 PH/02 Add the verb name to the "unknown ACL verb" error.
3454 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3456 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3458 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3459 directory not expanded when it should be if an expanded home directory
3460 was set for the address (which is overridden by the transport).
3462 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3465 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3466 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3467 because it is too late at that time, and has no effect.
3469 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3470 security issue with \' (bugzilla #107). I could not use the
3471 PQescapeStringConn() function, because it needs a PGconn value as one of
3474 PH/08 When testing addresses using -bt, indicate those final addresses that
3475 are duplicates that would not cause an additional delivery. At least one
3476 person was confused, thinking that -bt output corresponded to deliveries.
3477 (Suppressing duplicates isn't a good idea as you lose the information
3478 about possibly different redirections that led to the duplicates.)
3480 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3481 systems where poll() doesn't work, in particular OS X.
3483 PH/10 Added more information to debugging output for retry time not reached.
3485 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3486 operations in malware.c.
3488 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3491 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3492 syslog_duplication set false, log lines that would normally be written
3493 both the the main log and to the reject log were not written to syslog at
3496 PH/14 In the default configuration, change the use of "message" in ACL warn
3497 statements to "add_header".
3499 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3500 not followed by a command (e.g. "seen endif").
3502 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3503 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3506 PH/17 Added extra conditions to the default value of delay_warning_condition
3510 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3511 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3512 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3515 The Auto-Submitted: and various List- headers are standardised, whereas I
3516 don't think Precedence: ever was.
3518 PH/18 Refactored debugging code in route_finduser() to show more information,
3519 in particular, the error code if getpwnam() issues one.
3521 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3522 This is apparently needed in addition to the PH/07 change above to avoid
3523 any possible encoding problems.
3525 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3526 but not after initializing Perl.
3528 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3529 output them only if debugging. By default they are written stderr,
3530 apparently, which is not desirable.
3532 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3535 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3538 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3540 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3541 authenticated or an ident call has been made. Suppress the default
3542 values for $authenticated_id and $authenticated_sender (but permit -oMai
3543 and -oMas) when testing with -bh.
3545 PH/24 Re-jigged the order of the tests in the default configuration so that the
3546 tests for valid domains and recipients precede the DNS black list and CSA
3547 tests, on the grounds that those ones are more expensive.
3549 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3550 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3551 This bug exists in every version of Exim that I still have, right back to
3554 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3555 However, an attempt to turn on case-sensitivity in a regex key by
3556 including (?-i) didn't work because the subject string was already
3557 lowercased, and the effects were non-intuitive. It turns out that a
3558 one-line patch can be used to allow (?-i) to work as expected.
3564 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3565 other effects) broke the use of negated acl sub-conditions.
3567 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3570 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3571 "Deny" causes Exim to reject the incoming connection with a 554 error.
3572 Unfortunately, if there is a major crisis, such as a disk failure,
3573 tcp-wrappers gives "deny", whereas what one would like would be some
3574 kind of temporary error. A kludge has been added to help with this.
3575 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3576 554 error is used if errno is still zero or contains ENOENT (which occurs
3577 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3580 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3582 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3583 errors. Otherwise a message that provokes a temporary error (when other
3584 messages do not) can cause a whole host to time out.
3586 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3587 addresses were routed directly to files or pipes from a redirect router.
3588 File deliveries just didn't batch; pipe deliveries might have suffered
3591 PH/06 A failure to get a lock for a hints database would erroneously always say
3592 "Failed to get write lock", even when it was really a read lock.
3594 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3595 of 0600. This has been changed to use the value of the lockfile_mode
3596 option (which defaults to 0600).
3598 PH/08 Applied small patch from the Sieve maintainer.
3600 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3601 folder from quota calculations, a direct delivery into this folder messed
3602 up the contents of the maildirsize file. This was because the regex was
3603 used only to exclude .Trash (or whatever) when the size of the mailbox
3604 was calculated. There was no check that a delivery was happening into an
3605 excluded directory. This bug has been fixed by ignoring all quota
3606 processing for deliveries into excluded directories.
3608 PH/10 Added the maildirfolder_create_regex option to appendfile.
3614 PH/01 The code for finding all the local interface addresses on a FreeBSD
3615 system running IPv6 was broken. This may well have applied to all BSD
3616 systems, as well as to others that have similar system calls. The broken
3617 code found IPv4 interfaces correctly, but gave incorrect values for the
3618 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3619 that it would not match correctly against @[] and not recognize the IPv6
3622 PH/02 The ipliteral router was not recognizing addresses of the form user@
3623 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3625 PH/03 Added disable_ipv6.
3627 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3628 decoded form, because it is most often used to construct To: headers
3629 lines in autoreplies, and the decoded form may well be syntactically
3630 invalid. However, $reply_address has leading white space removed, and all
3631 newlines turned into spaces so that the autoreply transport does not
3634 PH/05 If group was specified without a user on a router, and no group or user
3635 was specified on a transport, the group from the router was ignored.
3637 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3638 for visible compile-time settings that can be used to change these
3639 numbers, for those that want even more. Backwards compatibility with old
3640 spool files has been maintained. However, going back to a previous Exim
3641 release will lost any variables that are in spool files.
3643 PH/07 Two small changes when running in the test harness: increase delay when
3644 passing a TCP/IP connection to a new process, in case the original
3645 process has to generate a bounce, and remove special handling of
3646 127.0.0.2 (sic), which is no longer necessary.
3648 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3649 be the same on different OS.
3651 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3654 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3655 whether --show-vars was specified or not
3657 JJ/02 exipick: Added support for new ACL variable spool format introduced
3660 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3661 syntactically invalid From: or Reply-to: line, and a filter used this to
3662 generate an autoreply, and therefore failed to obtain an address for the
3663 autoreply, Exim could try to deliver to a non-existent relative file
3664 name, causing unrelated and misleading errors. What now happens is that
3665 it logs this as a hard delivery error, but does not attempt to create a
3668 PH/11 The exinext utility has a -C option for testing purposes, but although
3669 the given file was scanned by exinext itself; it wasn't being passed on
3670 when Exim was called.
3672 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3673 an end-of-file indication when reading a command response.
3675 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3676 compiled. In many other places in Exim, IPv6 addresses are always
3677 recognized, so I have changed this. It also means that IPv4 domain
3678 literals of the form [IPV4:n.n.n.n] are now always recognized.
3680 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3681 used if the router is not running as root, for example, when verifying at
3682 ACL time, or when using -bh. The debugging output from this situation was
3683 non-existent - all you got was a failure to exec. I have made two
3686 (a) Failures to set uid/gid, the current directory, or a process leader
3687 in a subprocess such as that created by queryprogram now generate
3688 suitable debugging output when -d is set.
3690 (b) The queryprogram router detects when it is not running as root,
3691 outputs suitable debugging information if -d is set, and then runs
3692 the subprocess without attempting to change uid/gid.
3694 PH/15 Minor change to Makefile for building test_host (undocumented testing
3697 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3698 additional section of a DNS packet that returns MX or SRV records.
3699 Instead, it always explicitly searches for A/AAAA records. This avoids
3700 major problems that occur when a DNS server includes only records of one
3701 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3702 fixed another bug: if SRV records were looked up and the corresponding
3703 address records were *not* found in the additional section, the port
3704 values from the SRV records were lost.
3706 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3707 using the correct key (the original address) when searching the retry
3708 rules in order to find which one to use for generating the retry hint.
3710 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3711 adding the default one. Similarly, if it contains a Reply-To: header, the
3712 errors_reply_to option, if set, is not used.
3714 PH/19 When calculating a retry time, Exim used to measure the "time since
3715 failure" by looking at the "first failed" field in the retry record. Now
3716 it does not use this if it is later than than the arrival time of the
3717 message. Instead it uses the arrival time. This makes for better
3718 behaviour in cases where some deliveries succeed, thus re-setting the
3719 "first failed" field. An example is a quota failure for a huge message
3720 when small messages continue to be delivered. Without this change, the
3721 "time since failure" will always be short, possible causing more frequent
3722 delivery attempts for the huge message than are intended.
3723 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3725 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3726 $1, $2, $3) because the numerical variables can be reset during some
3727 expansion items (e.g. "match"), thereby losing the authentication data.
3729 PH/21 Make -bV show the size of off_t variables so that the test suite can
3730 decide whether to run tests for quotas > 2G.
3732 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3733 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3734 filecount value is greater than 2G or if a quota value is greater than 2G
3735 on a system where the size of off_t is not greater than 4, a panic error
3738 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3739 never match. The debug and -bh output now contains an explicit error
3740 message indicating a malformed IPv4 address or mask.
3742 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3743 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3744 PH/23 above applies.
3746 PH/25 Do not write to syslog when running in the test harness. The only
3747 occasion when this arises is a failure to open the main or panic logs
3748 (for which there is an explicit test).
3750 PH/26 Added the /no_tell option to "control=freeze".
3752 PH/27 If a host name lookup failed very early in a connection, for example, if
3753 the IP address matched host_lookup and the reverse lookup yielded a name
3754 that did not have a forward lookup, an error message of the form "no IP
3755 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3756 could be logged. Now it outputs the IP address instead of "NULL".
3758 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3759 allows the sender and the authenticated sender to be set when
3760 submitting a message from within Exim. Since child_open_exim() is
3761 documented for local_scan(), the new function should be too.
3763 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3764 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3765 results in an empty string is now treated as unset.
3767 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3769 PH/31 Added sender_verify_fail logging option.
3771 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3772 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3773 tidied the source and removed it altogether.
3775 PH/33 When a queue run was abandoned because the load average was too high, a
3776 log line was always written; now it is written only if the queue_run log
3777 selector is set. In addition, the log line for abandonment now contains
3778 information about the queue run such as the pid. This is always present
3779 in "start" and "stop" lines but was omitted from the "abandon" line.
3781 PH/34 Omit spaces between a header name and the colon in the error message that
3782 is given when verify = headers_syntax fails (if there are lots of them,
3783 the message gets confusing).
3785 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3786 names, as there are now some PTR records that contain slashes. This check
3787 is only to protect against broken name servers that fall over on strange
3788 characters, so the fact that it applies to all lookups doesn't matter.
3790 PH/36 Now that the new test suite is complete, we can remove some of the
3791 special code in Exim that was needed for the old test suite. For example,
3792 sorting DNS records because real resolvers return them in an arbitrary
3793 order. The new test suite's fake resolver always returns records in the
3796 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3797 bounces) except when queue_only is set, to avoid logging races between
3798 the different processes.
3800 PH/38 Panic-die if .include specifies a non-absolute path.
3802 PH/39 A tweak to the "H" retry rule from its user.
3804 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3805 a label. They prevented compilation on older perls.
3807 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3808 a warning to be raised on newish perls.
3810 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3811 on queue. Changes to match documented behaviour of showing count of
3812 messages matching specified criteria.
3814 PH/40 Changed the default ident timeout from 30s to 5s.
3816 PH/41 Added support for the use of login_cap features, on those BSD systems
3817 that have them, for controlling the resources used by pipe deliveries.
3819 PH/42 The content-scanning code uses fopen() to create files in which to put
3820 message data. Previously it was not paying any attention to the mode of
3821 the files. Exim runs with umask(0) because the rest of the code creates
3822 files with open(), and sets the required mode explicitly. Thus, these
3823 files were ending up world-writeable. This was not a big issue, because,
3824 being within the spool directory, they were not world-accessible. I have
3825 created a function called modefopen, which takes an additional mode
3826 argument. It sets umask(777), creates the file, chmods it to the required
3827 mode, then resets the umask. All the relevant calls to fopen() in the
3828 content scanning code have been changed to use this function.
3830 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3831 to 24 hours. This avoids potential overflow problems when processing G
3832 and H retry rules. I suspect nobody ever tinkers with this value.
3834 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3836 PH/45 When the plaintext authenticator is running as a client, the server's
3837 challenges are checked to ensure they are valid base64 strings. By
3838 default, the authentication attempt is cancelled if an invalid string is
3839 received. Setting client_ignore_invalid_base64 true ignores these errors.
3840 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3841 they are received. Thus, the responses can be made to depend on the
3842 challenges. If an invalid string is ignored, an empty string is placed in
3845 PH/46 Messages that are created by the autoreply transport now contains a
3846 References: header, in accordance with RFCs 2822 and 3834.
3848 PH/47 Added authenticated_sender_force to the smtp transport.
3850 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3852 PH/49 Installed latest patch from the Sieve maintainer.
3854 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3855 was also set, the appendfile transport was unnecessarily scanning a
3856 directory of message files (e.g. for maildir delivery) to find the count
3857 of files (along with the size), even though it did not need this
3858 information. It now does the scan only if it needs to find either the
3859 size of the count of files.
3861 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3863 PH/52 Two bugs concerned with error handling when the smtp transport is
3866 (i) Exim was not creating retry information for temporary errors given
3867 for individual recipients after the DATA command when the smtp transport
3868 was used in LMTP mode. This meant that they could be retried too
3869 frequently, and not timed out correctly.
3871 (ii) Exim was setting the flag that allows error details to be returned
3872 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3873 recipients that were returned after the DATA command.
3875 PH/53 This is related to PH/52, but is more general: for any failing address,
3876 when detailed error information was permitted to be returned to the
3877 sender, but the error was temporary, then after the final timeout, only
3878 "retry timeout exceeded" was returned. Now it returns the full error as
3879 well as "retry timeout exceeded".
3881 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3882 do this, and (what is worse) MTAs that accept it.
3884 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3885 will now be deprecated.
3887 PH/56 New os.c-cygwin from the Cygwin maintainer.
3889 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3890 formats (previously only available in exim formats via -bpr, -bpru,
3891 and -bpra. Now also available in native and exiqgrep formats)
3893 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3894 with very large, slow to parse queues
3896 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3898 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3900 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3901 responses to authentication challenges, though it was showing the
3902 challenges; (ii) I've removed the CR characters from the debug output for
3905 PH/58 Allow for the insertion of a newline as well as a space when a string
3906 is turned into more than one encoded-word during RFC 2047 encoding. The
3907 Sieve code now uses this.
3909 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3910 data_4xx, lost_connection, tls_required.
3912 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3913 message was being sent as an SMTP response.
3915 PH/61 Add -l and -k options to exicyclog.
3917 PH/62 When verifying, if an address was redirected to one new address, so that
3918 verification continued, and the new address failed or deferred after
3919 having set something in $address_data, the value of $address_data was not
3920 passed back to the ACL. This was different to the case when no
3921 redirection occurred. The value is now passed back in both cases.
3923 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3924 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3925 use login_cap.h, so on its own it isn't the distinguishing feature. The
3926 new name refers directly to the setclassresources() function.
3928 PH/65 Added configuration files for NetBSD3.
3930 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3932 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3933 is preferred over IPv4.
3935 PH/68 The bounce_return_message and bounce_return_body options were not being
3936 honoured for bounces generated during the reception of non-SMTP messages.
3937 In particular, this applied to messages rejected by the ACL. This bug has
3938 been fixed. However, if bounce_return_message is true and bounce_return_
3939 body is false, the headers that are returned for a non-SMTP message
3940 include only those that have been read before the error was detected.
3941 (In the case of an ACL rejection, they have all been read.)
3943 PH/69 The HTML version of the specification is now built in a directory called
3944 spec_html instead of spec.html, because the latter looks like a path with
3945 a MIME-type, and this confuses some software.
3947 PH/70 Catch two compiler warnings in sieve.c.
3949 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3950 function verify_get_ident() calls ip_connect() to connect a socket, but
3951 if the "connect()" function timed out, ip_connect() used to close the
3952 socket. However, verify_get_ident() also closes the socket later, and in
3953 between Exim writes to the log, which may get opened at this point. When
3954 the socket was closed in ip_connect(), the log could get the same file
3955 descriptor number as the socket. This naturally causes chaos. The fix is
3956 not to close the socket in ip_connect(); the socket should be closed by
3957 the function that creates it. There was only one place in the code where
3958 this was missing, in the iplookup router, which I don't think anybody now
3959 uses, but I've fixed it anyway.
3961 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3962 well as to direct DNS lookups. Otherwise the handling of names in host
3963 lists is inconsistent and therefore confusing.
3969 PH/01 Two changes to the default runtime configuration:
3971 (1) Move the checks for relay_from_hosts and authenticated clients from
3972 after to before the (commented out) DNS black list checks.
3974 (2) Add control=submission to the relay_from_hosts and authenticated
3975 clients checks, on the grounds that messages accepted by these
3976 statements are most likely to be submissions.
3978 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3980 (1) Generate an error if the third argument for the ${prvs expansion is
3983 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3986 (3) Reset the variables that are obtained from the first argument of
3987 ${prvscheck and used in the second argument before leaving the code,
3988 because their memory is reclaimed, so using them afterwards may do
3991 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3992 one (it's much easier than Tom thought :-).
3994 (5) Because of (4), we can now allow for the use of $prvscheck_result
3995 inside the third argument.
3997 PH/03 For some reason, the default setting of PATH when running a command from
3998 a pipe transport was just "/usr/bin". I have changed it to
4001 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4002 anything to be listed in the output from -bV.
4004 PH/05 When a filter generated an autoreply, the entire To: header line was
4005 quoted in the delivery log line, like this:
4007 => >A.N.Other <ano@some.domain> <original@ddress> ...
4009 This has been changed so that it extracts the operative address. There
4010 may be more than one such address. If so, they are comma-separated, like
4013 => >ano@some.domain,ona@other.domain <original@ddress> ...
4015 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4016 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4017 not being looked up in the rDNS to get a host name, Exim was showing the
4018 IP address twice in Received: lines, even though the IP addresses were
4019 identical. For example:
4021 Received: from [1.2.3.4] (helo=[1.2.3.4])
4023 However, if the real host name was known, it was omitting the HELO data
4024 if it matched the actual IP address. This has been tidied up so that it
4025 doesn't show the same IP address twice.
4027 PH/07 When both +timestamp and +memory debugging was on, the value given by
4028 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4029 called by the debug printing, thereby overwriting the timestamp buffer.
4030 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4032 PH/08 When the original message was included in an autoreply transport, it
4033 always said "this is a copy of the message, including all the headers",
4034 even if body_only or headers_only was set. It now gives an appropriate
4037 PH/09 Applied a patch from the Sieve maintainer which:
4039 o fixes some comments
4040 o adds the (disabled) notify extension core
4041 o adds some debug output for the result of if/elsif tests
4042 o points to the current vacation draft in the documentation
4043 and documents the missing references header update
4047 o fixes a bug in processing the envelope test (when testing
4048 multiple envelope elements, the last element determined the
4051 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4052 Electronic Mail") by including:
4054 Auto-submitted: auto-generated
4056 in the messages that it generates (bounce messages and others, such as
4057 warnings). In the case of bounce messages for non-SMTP messages, there was
4058 also a typo: it was using "Auto_submitted" (underscore instead of
4059 hyphen). Since every message generated by Exim is necessarily in response
4060 to another message, thes have all been changed to:
4062 Auto-Submitted: auto-replied
4064 in accordance with these statements in the RFC:
4066 The auto-replied keyword:
4068 - SHOULD be used on messages sent in direct response to another
4069 message by an automatic process,
4071 - MUST NOT be used on manually-generated messages,
4073 - MAY be used on Delivery Status Notifications (DSNs) and Message
4074 Disposition Notifications (MDNs),
4076 - MUST NOT be used on messages generated by automatic or periodic
4077 processes, except for messages which are automatic responses to
4080 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4081 to the default Received: header definition.
4083 PH/12 Added log selector acl_warn_skipped (default on).
4085 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4086 variables because (a) they are in the wrong storage pool and (b) even if
4087 they were copied, it wouldn't work properly because of the caching.
4089 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4090 checking when decoding. Apparently there are clients that generate
4091 overlong encoded strings. Why am I not surprised?
4093 PH/15 If the first argument of "${if match_address" was not empty, but did not
4094 contain an "@" character, Exim crashed. Now it writes a panic log message
4095 and treats the condition as false.
4097 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4099 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4100 extension "envelope-auth". The code is finished and in agreement with
4101 other implementations, but there is no documentation so far and in fact,
4102 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4103 not changing the active code.
4105 Print executed "if" and "elsif" statements when debugging is used. This
4106 helps a great deal to understand what a filter does.
4108 Document more things not specified clearly in RFC3028. I had all this
4109 sorted out, when out of a sudden new issues came to my mind. Oops."
4111 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4114 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4115 canonical form (as documented). However, after a host name lookup from
4116 the IP address, check_host() was doing a simple string comparison with
4117 addresses acquired from the DNS when checking that the found name did
4118 have the original IP as one of its addresses. Since any found IPv6
4119 addresses are likely to be in abbreviated form, the comparison could
4120 fail. Luckily, there already exists a function for doing the comparison
4121 by converting both addresses to binary, so now that is used instead of
4122 the text comparison.
4124 PH/20 There was another similar case to PH/19, when a complete host name was
4125 given in a host list; looking up its IP address could give an abbreviated
4126 form, whereas the current host's name might or might not be abbreviated.
4127 The same fix has been applied.
4133 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4134 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4137 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4138 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4140 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4142 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4143 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4144 Also, for "Auto-Submitted", treat anything other than "no" as
4145 non-personal, in accordance with RFC 3834. (Previously it treated
4146 anything starting "auto-" as non-personal.)
4148 TF/01 The control=submission/name=... option had a problem with syntax
4149 errors if the name included a slash character. The /name= option
4150 now slurps the rest of the string, so it can include any characters
4151 but it must come last in the list of options (after /sender_retain
4154 PH/05 Some modifications to the interface to the fake nameserver for the new
4162 TK/01 Added the "success_on_redirect" address verification option. See
4163 NewStuff for rationale and an example.
4165 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4167 PH/02 Patch to exigrep to allow it to work on syslog lines.
4169 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4170 fread() to skip over the body file's header line, because in Cygwin the
4171 header line is locked and is inaccessible.
4173 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4174 co-exist for some time) to make it clear that it is the Exim ID that is
4175 referenced, not the Message-ID: header line.
4177 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4178 string_format() function, because snprintf() does not exist on all
4181 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4182 ${stat: expansion item.
4184 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4185 protocol synchronization error", to keep the pedants happy.
4187 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4188 well as for IRIX systems, when gcc is being used. See the host.c source
4191 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4193 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4196 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4197 transport and to the smtp transport in LMTP mode.
4199 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4201 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4202 run a filter in a subprocess. This could lead to confusion in subsequent
4203 lookups in the parent process. There should also be a search_tidyup() at
4204 the end of the subprocess.
4206 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4207 only if the host matched helo_try_verify_hosts, which caused the
4208 verification to occur when the EHLO/HELO command was issued. The ACL just
4209 tested the remembered result. Now, if a previous verification attempt has
4210 not happened, "verify = helo" does it there and then.
4212 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4214 TK/03 Fix log output including CR from clamd.
4216 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4217 exist provoked a memory error which could cause a segfault.
4219 PH/15 Installed PCRE 6.2
4221 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4223 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4224 of the problem. Specifically, suggested +O2 rather than +O1 for the
4227 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4229 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4230 record, the port was not correctly logged when the outgoing_port log
4231 selector was set (it logged the transort's default port).
4233 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4234 fallback_hosts, and "hosts" in the smtp transport.
4236 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4237 host errors such as "Connection refused".
4239 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4240 authentication with radiusclient 0.4.9:
4242 - Error returned from rc_read_config was caught wrongly
4243 - Username/password not passed on to radius server due to wrong length.
4245 The presumption is that some radiusclient API changes for 4.51/PH/17
4246 were not taken care of correctly. The code is still untested by me (my
4247 Linux distribution still has 0.3.2 of radiusclient), but it was
4248 contributed by a Radius user.
4250 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4251 expanding the "port" option of the smtp transport.
4253 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4254 while reading a MIME header. Thanks to Tom Hughes for a patch.
4256 PH/24 Include config.h inside local_scan.h so that configuration settings are
4259 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4260 that in an ACL for RCPT (for example), you can examine exactly what was
4263 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4264 commands, but it was not correctly comparing the address with the actual
4265 client host address. Thus, it would show the EHLO address in Received:
4266 header lines when this was not necessary.
4268 PH/27 Added the % operator to ${eval:}.
4270 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4271 it should be ignoring failures (because with -C, for example, it has lost
4272 privilege). It wasn't ignoring creation failures other than "already
4275 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4278 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4279 because an input line was too long, either on its own, or by virtue of
4280 too many continuations, the temporary file was not being removed, and the
4281 return code was incorrect.
4283 PH/31 Missing "BOOL" in function definition in filtertest.c.
4285 PH/32 Applied Sieve patches from the maintainer.
4287 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4289 PH/33 Added "verify = not_blind".
4291 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4292 Local/Makefile (with some defaults set). These are used in built scripts
4293 such as exicyclog, but they have never been used in the exim_install
4294 script (though there are many overriding facilities there). I have
4295 arranged that the exim_install script now takes note of these two
4298 PH/35 Installed configuration files for Dragonfly.
4300 PH/36 When a locally submitted message by a trusted user did not contain a
4301 From: header, and the sender address was obtained from -f or from an SMTP
4302 MAIL command, and the trusted user did not use -F to supply a sender
4303 name, $originator_name was incorrectly used when constructing a From:
4304 header. Furthermore, $originator_name was used for submission mode
4305 messages from external hosts without From: headers in a similar way,
4306 which is clearly wrong.
4308 PH/37 Added control=suppress_local_fixups.
4310 PH/38 When log_selector = +received_sender was set, and the addition of the
4311 sender made the log line's construction buffer exactly full, or one byte
4312 less than full, an overflow happened when the terminating "\n" was
4315 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4316 when the result of a list match is failure because a DNS lookup failed.
4318 PH/40 RM_COMMAND is now used in the building process.
4320 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4321 the "build-* directories that it finds.
4323 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4324 address in a domain literal was a prefix of an interface address.
4326 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4327 when verifying a sender address, unless rewrite_headers is false.
4329 PH/44 Wrote a long comment about why errors_to addresses are verified as
4330 recipients, not senders.
4332 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4333 the ratelimit ACL was added.
4335 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4337 PH/46 Added extra information about PostgreSQL errors to the error string.
4339 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4340 avoiding the need to install special zones in a real server. This is
4341 backwards compatible; if it can't find the fake resolver, it drops back.
4342 Thus, both old and new test suites can be run.
4344 TF/02 Added util/ratelimit.pl
4346 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4347 clock is set back in time.
4349 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4350 Candler <B.Candler@pobox.com>.
4352 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4353 OK for addresses that are the result of redirections.
4355 PH/48 A number of further additions for the benefit of the new test suite,
4356 including a fake gethostbyname() that interfaces to the fake DNS resolver
4359 TF/06 The fix for widen_domains has also been applied to qualify_single and
4360 search_parents which are the other dnslookup options that can cause
4363 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4366 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4368 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4369 service" if it encountered an MX record with an empty target hostname.
4370 The message is now "an MX or SRV record indicated no SMTP service".
4372 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4373 if the DNS of the sending site is misconfigured. This is quite a
4374 common situation. This change restores the behaviour of treating a
4375 helo verification defer as a failure.
4377 PH/51 If self=fail was set on a router, the bounce message did not include the
4378 actual error message.
4384 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4386 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4387 command itself ended in error, the underlying message about the transport
4388 filter timeout was being overwritten with the pipe command error. Now the
4389 underlying error message should be appended to the second error message.
4391 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4393 PH/02 Reduce the amount of output that "make" produces by default. Full output
4394 can still be requested.
4396 PH/03 The warning log line about a condition test deferring for a "warn" verb
4397 was being output only once per connection, rather than after each
4398 occurrence (because it was using the same function as for successful
4399 "warn" verbs). This seems wrong, so I have changed it.
4401 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4402 it should not have, which might have caused a crash in the right
4403 circumstances, but probably never did.
4405 PH/04 Installed a modified version of Tony Finch's patch to make submission
4406 mode fix the return path as well as the Sender: header line, and to
4407 add a /name= option so that you can make the user's friendly name appear
4410 TF/03 Added the control = fakedefer ACL modifier.
4412 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4413 Mark Lowes for thorough testing.
4415 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4417 TK/03 Merged latest SRS patch from Miles Wilton.
4419 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4420 with the definition in sysexits.h (which is #included earlier).
4421 Fortunately, Exim does not actually use EX_OK. The code used to try to
4422 preserve the sysexits.h value, by assuming that macro definitions were
4423 scanned for macro replacements. I have been disabused of this notion,
4424 so now the code just undefines EX_OK before #including unistd.h.
4426 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4427 in the smtp transport. When a block could not be written in a single
4428 write() function, the timeout was being re-applied to each part-write.
4429 This seems wrong - if the receiver was accepting one byte at a time it
4430 would take for ever. The timeout is now adjusted when this happens. It
4431 doesn't have to be particularly precise.
4433 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4434 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4436 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4437 without a domain if the check to <postmaster@domain> fails.
4439 SC/01 Eximstats: added -xls and the ability to specify output files
4440 (patch written by Frank Heydlauf).
4442 SC/02 Eximstats: use FileHandles for outputting results.
4444 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4446 SC/04 Eximstats: fixed display of large numbers with -nvr option
4448 SC/05 Eximstats: fixed merging of reports with empty tables.
4450 SC/06 Eximstats: added the -include_original_destination flag
4452 SC/07 Eximstats: removed tabs and trailing whitespace.
4454 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4456 TK/06 MBOX spool code: Add real "From " MBOX separator line
4457 so the .eml file is really in mbox format (even though
4458 most programs do not really care). Patch from Alex Miller.
4460 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4461 The latter is generated from $received_to and is only set if the
4462 message has one envelope recipient. SA can use these headers,
4463 obviously out-of-the-box. Patch from Alex Miller.
4465 PH/08 The ${def test on a variable was returning false if the variable's
4466 value was "0", contrary to what the specification has always said!
4467 The result should be true unless the variable is empty.
4469 PH/09 The syntax error of a character other than { following "${if
4470 def:variable_name" (after optional whitespace) was not being diagnosed.
4471 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4472 accidental colon was present, for example, could give incorrect results.
4474 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4475 result is used (not including appendfile, where other changes are about
4478 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4479 This involved changing a lot of size variables from int to off_t. It
4480 should work with maildirs and everything.
4482 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4483 spamd dying while we are connected to it.
4485 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4488 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4489 function for BDB 4.3.
4491 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4493 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4494 the sources for PCRE have been re-arranged and more files are now
4497 PH/15 The code I had for printing potentially long long variables in PH/11
4498 above was not the best (it lost precision). The length of off_t variables
4499 is now inspected at build time, and an appropriate printing format (%ld
4500 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4501 to be "long long int" or "long int". This is needed for the internal
4502 formatting function string_vformat().
4504 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4505 the configuration file to be ":syslog", then the script "guesses" where
4506 the logs files are, rather than using the compiled in default. In our
4507 case the guess is not the same as the compiled default, so the script
4508 suddenly stopped working when I started to use syslog. The patch checks
4509 to see if log_file_path is "". If so, it attempts to read it from exim
4510 with no configuration file to get the compiled in version, before it
4511 falls back to the previous guessing code."
4513 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4514 implementing BATV in an Exim configuration. See NewStuff for the gory
4517 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4518 Makefile that are specific to HP-UX.
4520 PH/18 If the "use_postmaster" option was set for a recipient callout together
4521 with the "random" option, the postmaster address was used as the MAIL
4522 FROM address for the random test, but not for the subsequent recipient
4523 test. It is now used for both.
4525 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4526 patch removes a few documentation additions to RFC 3028, because the
4527 latest draft now contains them. It adds the new en;ascii-case comparator
4528 and a new error check for 8bit text in MIME parts. Comparator and
4529 require names are now matched exactly. I enabled the subaddress
4530 extension, but it is not well tested yet (read: it works for me)."
4532 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4533 rework some of the code of TK/09 above to avoid the hardwired use of
4534 "%lld" and "long long". Replaced the call to snprintf() with a call to
4537 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4538 records point to non-existent hosts", "retry timeout exceeded", and
4539 "retry time not reached for any host after a long failure period".
4541 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4542 experimental DomainKeys support:
4544 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4545 (2) On an error such as an illegally used "control", the wrong name for
4546 the control was given.
4548 These problems did NOT occur unless DomainKeys support was compiled.
4550 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4552 PH/24 Added ${if match_ip condition.
4554 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4555 databases so that it will be absolutely obvious if a crash occurs in the
4556 DB library. This is a regular occurrence (often caused by mis-matched
4559 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4560 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4561 picked up on a user's system that detects such things. There doesn't seem
4562 to be a gcc warning option for this - only an attribute that has to be
4563 put on the function's prototype. It seems that in Fedora Core 4 they have
4564 set this on a number of new functions. No doubt there will be more in due
4567 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4568 specify a transport. However, if an address that was verified by such a
4569 router was the subject of a callout, Exim crashed because it tried to
4570 read the rcpt_include_affixes from the non-existent transport. Now it
4571 just assumes that the setting of that option is false. This bug was
4572 introduced by 4.51/PH/31.
4574 PH/28 Changed -d+all to exclude +memory, because that information is very
4575 rarely of interest, but it makes the output a lot bigger. People tend to
4576 do -d+all out of habit.
4578 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4579 code in os-type was giving problems when libc.so lives in lib64, like on
4582 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4583 aren't the modern standard, and it seems that some systems' include files
4584 don't always have them. Exim was already checking for some of the newer
4585 ones like T_AAAA, and defining it itself. I've added checks for all the
4586 record types that Exim uses.
4588 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4589 not automatically generating a new one, as it is supposed to. This
4590 prevented TLS from working. If the file did exist, but contained invalid
4591 data, a new version was generated, as expected. It was only the case of a
4592 non-existent file that was broken.
4594 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4595 with a change in libdomainkeys > 0.64.
4597 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4598 from DNS. If the selector record carries the flag, it now has
4599 precedence over the domain-wide flag.
4601 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4603 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4604 the use of a port name that isn't defined in /etc/services) occurred, the
4605 message was deferred as in a normal delivery, and thus remained on the
4606 spool, instead of being failed because of the mua_wrapper setting. This
4607 is now fixed, and I tidied up some of the mua_wrapper messages at the
4610 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4611 the messages in a hash of arrays rather than using individual hashes.
4612 This is a bit cleaner and results in dramatic memory savings, albeit
4613 at a slight CPU cost.
4615 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4616 as requested by Marc Sherman.
4618 SC/10 Eximstats: added histograms for user specified patterns as requested
4621 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4623 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4624 fopen() in the content-scanning modules that did not already have it.
4630 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4631 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4633 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4635 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4637 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4638 to test Sieve filters that use "vacation".
4640 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4641 that changes the way the GnuTLS parameters are stored in the cache file.
4642 The new format can be generated externally. For backward compatibility,
4643 if the data in the cache doesn't make sense, Exim assumes it has read an
4644 old-format file, and it generates new data and writes a new file. This
4645 means that you can't go back to an older release without removing the
4648 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4649 work if there are any delivery delays because "one_time" forces the
4650 parent to be marked "delivered", so its unseen clone is never tried
4651 again. For this reason, Exim now forbids the simultaneous setting of
4654 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4655 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4656 entry if you want to know the details. The fix, however, neglected to
4657 consider the case when local delivery batching is involved. The test for
4658 "previously delivered" was not happening when checking to see if an
4659 address could be batched with a previous (undelivered) one; under
4660 certain circumstances this could lead to multiple deliveries to the same
4663 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4664 in its include files, and this causes problems building Exim.
4666 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4667 header_syntax) but Exim was just ignoring anything given after a slash.
4668 In particular, this caused confusion with an attempt to use "verify =
4669 reverse_host_lookup/defer_ok". An error is now given when options are
4670 supplied for verify items that do not have them. (Maybe reverse_host_
4671 lookup should have a defer_ok option, but that's a different point.)
4673 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4674 defined by RFC 821) to 2048, because there were problems with some AUTH
4675 commands, and RFC 1869 says the size should be increased for extended
4676 SMTP commands that take arguments.
4678 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4681 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4682 "unknown" error; now it says that the functionality isn't in the binary.
4684 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4685 an address' error message when a string expansion fails (syntax or
4686 whatever). Otherwise the password may appear in the log. Following change
4687 PH/42 below, there is no longer a chance of it appearing in a bounce
4690 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4692 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4693 of its addresses was ever tried. (Bugzilla bug #2.)
4695 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4696 the result incorrectly in the debug output. (It correctly added a newline
4697 to what was transported.)
4699 TF/01 Added $received_time.
4701 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4702 commented out examples of how to interface to a virus scanner and to
4703 SpamAssassin. Also added commented examples of av_scanner and
4704 spamd_address settings.
4706 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4707 and controls are allowed in which ACLs. There were a couple of minor
4708 errors. Some of the entries in the conditions table (which is a table of
4709 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4710 negation of where the condition IS allowed.
4712 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4714 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4715 header file does not have a version number, so I've had to invent a new
4716 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4717 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4718 radiusclient), but it was contributed by a Radius user.
4720 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4721 files or directories, for OpenSSL.
4723 PH/19 When an Exim process that is running as root has to create an Exim log
4724 file, it does so in a subprocess that runs as exim:exim so as to get the
4725 ownership right at creation (otherwise, other Exim processes might see
4726 the file with the wrong ownership). There was no test for failure of this
4727 fork() call, which would lead to the process getting stuck as it waited
4728 for a non-existent subprocess. Forks do occasionally fail when resources
4729 run out. I reviewed all the other calls to fork(); they all seem to check
4732 PH/20 When checking for unexpected SMTP input at connect time (before writing
4733 the banner), Exim was not dealing correctly with a non-positive return
4734 from the read() function. If the client had disconnected by this time,
4735 the result was a log entry for a synchronization error with an empty
4736 string after "input=" when read() returned zero. If read() returned -1
4737 (an event I could not check), uninitialized data bytes were printed.
4738 There were reports of junk text (parts of files, etc) appearing after
4741 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4743 PH/22 Added support for macro redefinition, and (re)definition in between
4744 driver and ACL definitions.
4746 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4747 forgetting to use the resulting value; it was using the unexpanded value.
4749 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4750 hadn't been configured. The fix is from Juergen Kreileder, who
4751 understands it better than I do:
4753 "Here's what I see happening with three configured cyrus_sasl
4754 authenticators configured (plain, login, cram-md5):
4756 On startup auth_cyrus_sasl_init() gets called for each of these.
4757 This means three calls to sasl_listmech() without a specified mech_list.
4758 => SASL tests which mechs of all available mechs actually work
4759 => three warnings about OTP not working
4760 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4762 With the patch, sasl_listmech() also gets called three times. But now
4763 SASL's mech_list option is set to the server_mech specified in the the
4764 authenticator. Or in other words, the answer from sasl_listmech()
4765 gets limited to just the mech you're testing for (which is different
4767 => the return list contains just 'plain' or 'login', 'cram-md5' or
4768 nothing depending on the value of ob->server_mech.
4770 I've just tested the patch: Authentication still works fine,
4771 unavailable mechs specified in the exim configuration are still
4772 caught, and the auth.log warnings about OTP are gone."
4774 PH/25 When debugging is enabled, the contents of the command line are added
4775 to the debugging output, even when log_selector=+arguments is not
4778 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4779 answer is "GNU", and only if the return is "GNU/something" is the answer
4782 PH/27 $acl_verify_message is now set immediately after the failure of a
4783 verification in an ACL, and so is available in subsequent modifiers. In
4784 particular, the message can be preserved by coding like this:
4786 warn !verify = sender
4787 set acl_m0 = $acl_verify_message
4789 Previously, $acl_verify_message was set only while expanding "message"
4790 and "log_message" when a very denied access.
4792 PH/28 Modified OS/os.c-Linux with
4794 -#ifndef OS_LOAD_AVERAGE
4795 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4797 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4798 nomenclature these days.)
4800 PH/29 Installed patch from the Sieve maintainer that adds the options
4801 sieve_useraddress and sieve_subaddress to the redirect router.
4803 PH/30 In these circumstances:
4804 . Two addresses routed to the same list of hosts;
4805 . First host does not offer TLS;
4806 . First host accepts first address;
4807 . First host gives temporary error to second address;
4808 . Second host offers TLS and a TLS session is established;
4809 . Second host accepts second address.
4810 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4811 and peerdn, if requested) that were in fact used only for the second
4814 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4815 attention to any local part prefix or suffix that was matched by the
4816 router that accepted the address. It now behaves in the same way as it
4817 does for delivery: the affixes are removed from the local part unless
4818 rcpt_include_affixes is set on the transport.
4820 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4821 timeout during the DATA phase of an incoming message.
4823 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4824 applied a patch sanctioned by the Sieve maintainer.
4826 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4827 the uid or gid is negative. A case of a negative gid caused this to be
4828 noticed. The fix allows for either to be negative.
4830 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4831 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4834 PH/36 The change PH/12 above was broken. Fixed it.
4836 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4837 the grounds that routing the same address twice would always produce the
4838 same answer. This might have been true once, but it is certainly no
4839 longer true now. Routing a child address may depend on the previous
4840 routing that produced that child. Some complicated redirection strategies
4841 went wrong when messages had multiple recipients, and made Exim's
4842 behaviour dependent on the order in which the addresses were given.
4844 I have moved the duplicate checking until after the routing is complete.
4845 Exim scans the addresses that are assigned to local and remote
4846 transports, and removes any duplicates. This means that more work will be
4847 done, as duplicates will always all be routed, but duplicates are
4848 presumably rare, so I don't expect this is of any significance.
4850 For deliveries to pipes, files, and autoreplies, the duplicate checking
4851 still happens during the routing process, since they are not going to be
4854 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4855 It corrects a timeout issue with spamd. This is Ian's comment: "The
4856 background is that sometimes spamd either never reads data from a
4857 connection it has accepted, or it never writes response data. The exiscan
4858 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4859 blindly assumes that writes won't block so it may never time out."
4861 PH/39 Allow G after quota size as well as K and M.
4863 PH/40 The value set for $authenticated_id in an authenticator may not contain
4864 binary zeroes or newlines because the value is written to log lines and
4865 to spool files. There was no check on this. Now the value is run through
4866 the string_printing() function so that such characters are converted to
4867 printable escape sequences.
4869 PH/41 $message_linecount is a new variable that contains the total number of
4870 lines in the message. Compare $body_linecount, which is the count for the
4873 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4874 bounce and delay warning messages, except in certain special cases, which
4877 (a) An SMTP error message from a remote host;
4878 (b) A message specified in a :fail: redirection;
4879 (c) A message specified in a "fail" command in a system filter;
4880 (d) A message specified in a FAIL return from the queryprogram router;
4881 (e) A message specified by the cannot_route_message router option.
4883 In these cases only, Exim does include the error details in bounce and
4884 warning messages. There are also a few cases where bland messages such
4885 as "unrouteable address" or "local delivery error" are given.
4887 PH/43 $value is now also set for the "else" part of a ${run expansion.
4889 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4890 being worked on, but at least Exim now implements the latest version to
4893 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4894 process to complete was treated as a delivery failure, a timeout while
4895 writing the message to the pipe was logged, but erroneously treated as a
4896 successful delivery. Such timeouts include transport filter timeouts. For
4897 consistency with the overall process timeout, these timeouts are now
4898 treated as errors, giving rise to delivery failures by default. However,
4899 there is now a new Boolean option for the pipe transport called
4900 timeout_defer, which, if set TRUE, converts the failures into defers for
4901 both kinds of timeout. A transport filter timeout is now identified in
4904 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4905 systems where "make" and "gmake" are different, calling "gmake" at top
4906 level broke things. I've arranged for the value of $(MAKE) to be passed
4907 from the Makefile to this script so that it can call the same version of
4911 A note about Exim versions 4.44 and 4.50
4912 ----------------------------------------
4914 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4915 changes of various kinds. As a consequence, a big documentation update was
4916 needed. This delayed the release for rather longer than seemed good, especially
4917 in the light of a couple of (minor) security issues. Therefore, the changes
4918 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4919 release. So 4.44 and 4.50 are in effect two different branches that both start
4922 I have left the 4.50 change log unchanged; it contains all the changes since
4923 4.43. The change log for 4.44 is below; many of its items are identical to
4924 those for 4.50. This seems to be the most sensible way to preserve the
4925 historical information.
4931 1. Minor wording change to the doc/README.SIEVE file.
4933 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4934 computation of the current number of files was incorrect.
4936 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4937 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4940 4. Give more explanation in the error message when the command for a transport
4941 filter fails to execute.
4943 5. There are several places where Exim runs a non-Exim command in a
4944 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4945 was being done only for the command run by the queryprogram router. It is
4946 now done for all such subprocesses. The other cases are: ${run, transport
4947 filters, and the commands run by the lmtp and pipe transports.
4949 6. Added CONFIGURE_GROUP build-time option.
4951 7. Some older OS have a limit of 256 on the maximum number of file
4952 descriptors. Exim was using setrlimit() to set 1000 as a large value
4953 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4954 systems. I've change it so that if it can't get 1000, it tries for 256.
4956 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4957 was an oversight, and furthermore, ever since the addition of extra
4958 controls (e.g. 4.43/32), the checks on when to allow different forms of
4959 "control" were broken. There should now be diagnostics for all cases when a
4960 control that does not make sense is encountered.
4962 9. Added the /retain_sender option to "control=submission".
4964 10. $recipients is now available in the predata ACL (oversight).
4966 11. Tidy the search cache before the fork to do a delivery from a message
4967 received from the command line. Otherwise the child will trigger a lookup
4968 failure and thereby defer the delivery if it tries to use (for example) a
4969 cached ldap connection that the parent has called unbind on.
4971 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4972 of $address_data from the recipient verification was clobbered by the
4973 sender verification.
4975 13. The value of address_data from a sender verification is now available in
4976 $sender_address_data in subsequent conditions in the ACL statement.
4978 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4980 15. Added a new option "connect=<time>" to callout options, to set a different
4983 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4984 was its contents. (It was OK if the option was not defined at all.)
4986 17. A "Completed" log line is now written for messages that are removed from
4987 the spool by the -Mrm option.
4989 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4990 information about exactly what failed.
4992 19. Added -dd to debug only the daemon process.
4994 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4995 handles timeouts, both on the server side and network timeouts. Renamed the
4996 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4998 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4999 It is now set to "smtps".
5001 22. $host_address is now set to the target address during the checking of
5002 ignore_target_hosts.
5004 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5005 being passed; this would have caused $sender_host_name to have been used if
5006 matching the list had actually called for a host name (not very likely,
5007 since this list is usually IP addresses). A host name is now passed as
5010 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5011 code that specifies a non-restarting handler (typically sigaction() in
5012 modern systems) in an attempt to fix a rare and obscure crash bug.
5014 25. Narrowed the window for a race in the daemon that could cause it to ignore
5015 SIGCHLD signals. This is not a major problem, because they are used only to
5016 wake it up if nothing else does.
5018 26. A malformed maildirsize file could cause Exim to calculate negative values
5019 for the mailbox size or file count. Odd effects could occur as a result.
5020 The maildirsize information is now recalculated if the size or filecount
5023 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5024 support for a long time. Removed HAVE_SYS_VFS_H.
5026 28. Installed the latest version of exipick from John Jetmore.
5028 29. In an address list, if the pattern was not a regular expression, an empty
5029 subject address (from a bounce message) matched only if the pattern was an
5030 empty string. Non-empty patterns were not even tested. This was the wrong
5031 because it is perfectly reasonable to use an empty address as part of a
5032 database query. An empty address is now tested by patterns that are
5033 lookups. However, all the other forms of pattern expect the subject to
5034 contain a local part and a domain, and therefore, for them, an empty
5035 address still always fails if the pattern is not itself empty.
5037 30. Exim went into a mad DNS loop when attempting to do a callout where the
5038 host was specified on an smtp transport, and looking it up yielded more
5039 than one IP address.
5041 31. Re-factored the code for checking spool and log partition space into a
5042 function that finds that data and another that does the check. The former
5043 is then used to implement four new variables: $spool_space, $log_space,
5044 $spool_inodes, and $log_inodes.
5046 32. The RFC2047 encoding function was originally intended for short strings
5047 such as real names; it was not keeping to the 75-character limit for
5048 encoded words that the RFC imposes. It now respects the limit, and
5049 generates multiple encoded words if necessary. To be on the safe side, I
5050 have increased the buffer size for the ${rfc2047: expansion operator from
5053 33. It is now permitted to omit both strings after an "if" condition; if the
5054 condition is true, the result is "true". As before, when the second string
5055 is omitted, a false condition yields an empty string. This makes it less
5056 cumbersome to write custom ACL and router conditions.
5058 34. Failure to deliver a bounce message always caused it to be frozen, even if
5059 there was an errors_to setting on the router. The errors_to setting is now
5062 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5063 canonical form (fully expanded) before being placed in
5064 $sender_host_address.
5066 36. The table in the code that translates DNS record types into text (T_A to
5067 "A" for instance) was missing entries for NS and CNAME. It is just possible
5068 that this could have caused confusion if both these types were looked up
5069 for the same domain, because the text type is used as part of Exim's
5070 per-process caching. But the chance of anyone hitting this buglet seems
5073 37. The dnsdb lookup has been extended in a number of ways.
5075 (1) There is a new type, "zns", which walks up the domain tree until it
5076 finds some nameserver records. It should be used with care.
5078 (2) There is a new type, "mxh", which is like "mx" except that it returns
5079 just the host names, not the priorities.
5081 (3) It is now possible to give a list of domains (or IP addresses) to be
5082 looked up. The behaviour when one of the lookups defers can be
5083 controlled by a keyword.
5085 (4) It is now possible to specify the separator character for use when
5086 multiple records are returned.
5088 38. The dnslists ACL condition has been extended: it is now possible to supply
5089 a list of IP addresses and/or domains to be looked up in a particular DNS
5092 39. Added log_selector=+queue_time_overall.
5094 40. When running the queue in the test harness, wait just a tad after forking a
5095 delivery process, to get repeatability of debugging output.
5097 41. Include certificate and key file names in error message when GnuTLS fails
5098 to set them up, because the GnuTLS error message doesn't include the name
5099 of the failing file when there is a problem reading it.
5101 42. Allow both -bf and -bF in the same test run.
5103 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5105 44. The "Exiscan patch" is now merged into the mainline Exim source.
5107 45. Sometimes the final signoff response after QUIT could fail to get
5108 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5109 < 0 before doing a fflush(). This bug looks as though it goes back to the
5110 introduction of TLS in release 3.20, but "sometimes" must have been rare
5111 because the tests only now provoked it.
5113 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5114 (this can affect the format of dates).
5116 47. exim_tidydb, when checking for the continued existence of a message for
5117 which it has found a message-specific retry record, was not finding
5118 messages that were in split spool directories. Consequently, it was
5119 deleting retry records that should have stayed in existence.
5121 48. Steve fixed some bugs in eximstats.
5123 49. The SPA authentication driver was not abandoning authentication and moving
5124 on to the next authenticator when an expansion was forced to fail,
5125 contradicting the general specification for all authenticators. Instead it
5126 was generating a temporary error. It now behaves as specified.
5128 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5129 (the order specifies the preference for clients). The order is now AES256,
5130 AES128, 3DES, ARCFOUR128.
5132 51. Small patch to Sieve code - explicitly set From: when generating an
5135 52. Exim crashed if a remote delivery caused a very long error message to be
5136 recorded - for instance if somebody sent an entire SpamAssassin report back
5137 as a large number of 550 error lines. This bug was coincidentally fixed by
5138 increasing the size of one of Exim's internal buffers (big_buffer) that
5139 happened as part of the Exiscan merge. However, to be on the safe side, I
5140 have made the code more robust (and fixed the comments that describe what
5143 53. Now that there can be additional text after "Completed" in log lines (if
5144 the queue_time_overall log selector is set), a one-byte patch to exigrep
5145 was needed to allow it to recognize "Completed" as not the last thing in
5148 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5149 patch that reportedly fixes this has been added. I am not expert enough to
5150 create a test for it. This is what the patch creator wrote:
5152 "I found a little strange behaviour of ldap code when working with
5153 Windows 2003 AD Domain, where users was placed in more than one
5154 Organization Units. When I tried to give exim partial DN, the exit code
5155 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5156 But simultaneously result of request was absolutely normal ldap result,
5157 so I produce this patch..."
5159 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5160 REFERENCE, so I have modified the code to exclude the patch when that macro
5163 55. Some experimental protocols are using DNS PTR records for new purposes. The
5164 keys for these records are domain names, not reversed IP addresses. The
5165 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5166 leaves it alone. Component reversal etc. now happens only for IP addresses.
5169 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5171 57. Double the size of the debug message buffer (to 2048) so that more of very
5172 long debug lines gets shown.
5174 58. The exicyclog utility now does better if the number of log files to keep
5175 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5177 59. Two changes related to the smtp_active_hostname option:
5179 (1) $smtp_active_hostname is now available as a variable.
5180 (2) The default for smtp_banner uses $smtp_active_hostname instead
5181 of $primary_hostname.
5183 60. The host_aton() function is supposed to be passed a string that is known
5184 to be a valid IP address. However, in the case of IPv6 addresses, it was
5185 not checking this. This is a hostage to fortune. Exim now panics and dies
5186 if the condition is not met. A case was found where this could be provoked
5187 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5188 components; fortuitously, this particular loophole had already been fixed
5189 by change 4.50/55 above.
5191 If there are any other similar loopholes, the new check in host_aton()
5192 itself should stop them being exploited. The report I received stated that
5193 data on the command line could provoke the exploit when Exim was running as
5194 exim, but did not say which command line option was involved. All I could
5195 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5196 running as the user.
5199 61. There was a buffer overflow vulnerability in the SPA authentication code
5200 (which came originally from the Samba project). I have added a test to the
5201 spa_base64_to_bits() function which I hope fixes it.
5204 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5205 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5207 63. The daemon start-up calls getloadavg() while still root for those OS that
5208 need the first call to be done as root, but it missed one case: when
5209 deliver_queue_load_max is set with deliver_drop_privilege. This is
5210 necessary for the benefit of the queue runner, because there is no re-exec
5211 when deliver_drop_privilege is set.
5213 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5214 This has been fixed.
5216 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5217 was in use, was not putting the data itself into the right store pool;
5218 consequently, it could be overwritten for a subsequent message in the same
5219 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5222 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5224 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5225 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5226 the function were treating the return as a boolean value, which happened to
5227 work because 0=false and not-0=true, but is not correct code.
5229 68. The host_aton() function was not handling scoped IPv6 addresses (those
5230 with, for example, "%eth0" on the end) correctly.
5232 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5233 negated items (that is, ~something) in unsigned ints. Some compilers
5234 apparently mutter when there is no cast.
5236 70. If an address verification called from an ACL failed, and did not produce a
5237 user-specific message (i.e. there was only a "system" message), nothing was
5238 put in $acl_verify_message. In this situation, it now puts the system
5241 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5242 session; change 4.31/43 added the unwanted input to the log line - except
5243 that it did not do this in the start of session case. It now does.
5245 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5246 This is wrong and can cause the other end to generate a synchronization
5247 error if it is another Exim or anything else that does the synchronization
5248 check. A QUIT command is no longer sent after a timeout.
5250 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5251 during host lookups.
5253 74. The defer_ok option of callout verification was not working if it was used
5254 when verifying addresses in header lines, that is, for this case:
5256 verify = header_sender/callout=defer_ok
5258 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5259 those file descriptors could be used for SMTP connections. If anything
5260 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5261 could be sent to the SMTP client, causing chaos. The daemon now opens
5262 stdin, stdout, and stderr to /dev/null when it puts itself into the
5265 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5266 log by default. The user can override this with suitable Perl magic.
5268 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5269 the log message when discard triggers, was not working for the DATA ACL or
5270 for the non-SMTP ACL.
5272 78. Error message wording change in sieve.c.
5274 79. If smtp_accept_max_per_host was set, the number of connections could be
5275 restricted to fewer than expected, because the daemon was trying to set up
5276 a new connection before checking whether the processes handling previous
5277 connections had finished. The check for completed processes is now done
5278 earlier. On busy systems, this bug wouldn't be noticed because something
5279 else would have woken the daemon, and it would have reaped the completed
5282 80. If a message was submitted locally by a user whose login name contained one
5283 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5284 It caused a spool format error. I have fixed the spool reading code. A
5285 related problem was that the "from" clause in the Received: line became
5286 illegal because of the space(s). It is now covered by ${quote_local_part.
5288 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5291 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5293 83. Patch from Sieve maintainer to fix unterminated string problem in
5294 "vacation" handling.
5296 84. Some minor changes to the Linux configuration files to help with other
5297 OS variants using glibc.
5299 85. One more patch for Sieve to update vacation handling to latest spec.
5302 ----------------------------------------------------
5303 See the note above about the 4.44 and 4.50 releases.
5304 ----------------------------------------------------
5310 1. Change 4.43/35 introduced a bug that caused file counts to be
5311 incorrectly computed when quota_filecount was set in an appendfile
5314 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5315 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5318 3. Give more explanation in the error message when the command for a transport
5319 filter fails to execute.
5321 4. There are several places where Exim runs a non-Exim command in a
5322 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5323 was being done only for the command run by the queryprogram router. It is
5324 now done for all such subprocesses. The other cases are: ${run, transport
5325 filters, and the commands run by the lmtp and pipe transports.
5327 5. Some older OS have a limit of 256 on the maximum number of file
5328 descriptors. Exim was using setrlimit() to set 1000 as a large value
5329 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5330 systems. I've change it so that if it can't get 1000, it tries for 256.
5332 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5333 was an oversight, and furthermore, ever since the addition of extra
5334 controls (e.g. 4.43/32), the checks on when to allow different forms of
5335 "control" were broken. There should now be diagnostics for all cases when a
5336 control that does not make sense is encountered.
5338 7. $recipients is now available in the predata ACL (oversight).
5340 8. Tidy the search cache before the fork to do a delivery from a message
5341 received from the command line. Otherwise the child will trigger a lookup
5342 failure and thereby defer the delivery if it tries to use (for example) a
5343 cached ldap connection that the parent has called unbind on.
5345 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5346 of $address_data from the recipient verification was clobbered by the
5347 sender verification.
5349 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5350 was its contents. (It was OK if the option was not defined at all.)
5352 11. A "Completed" log line is now written for messages that are removed from
5353 the spool by the -Mrm option.
5355 12. $host_address is now set to the target address during the checking of
5356 ignore_target_hosts.
5358 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5359 being passed; this would have caused $sender_host_name to have been used if
5360 matching the list had actually called for a host name (not very likely,
5361 since this list is usually IP addresses). A host name is now passed as
5364 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5365 code that specifies a non-restarting handler (typically sigaction() in
5366 modern systems) in an attempt to fix a rare and obscure crash bug.
5368 15. Narrowed the window for a race in the daemon that could cause it to ignore
5369 SIGCHLD signals. This is not a major problem, because they are used only to
5370 wake it up if nothing else does.
5372 16. A malformed maildirsize file could cause Exim to calculate negative values
5373 for the mailbox size or file count. Odd effects could occur as a result.
5374 The maildirsize information is now recalculated if the size or filecount
5377 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5378 support for a long time. Removed HAVE_SYS_VFS_H.
5380 18. Updated exipick to current release from John Jetmore.
5382 19. Allow an empty sender to be matched against a lookup in an address list.
5383 Previously the only cases considered were a regular expression, or an
5386 20. Exim went into a mad DNS lookup loop when doing a callout where the
5387 host was specified on the transport, if the DNS lookup yielded more than
5390 21. The RFC2047 encoding function was originally intended for short strings
5391 such as real names; it was not keeping to the 75-character limit for
5392 encoded words that the RFC imposes. It now respects the limit, and
5393 generates multiple encoded words if necessary. To be on the safe side, I
5394 have increased the buffer size for the ${rfc2047: expansion operator from
5397 22. Failure to deliver a bounce message always caused it to be frozen, even if
5398 there was an errors_to setting on the router. The errors_to setting is now
5401 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5402 canonical form (fully expanded) before being placed in
5403 $sender_host_address.
5405 24. Updated eximstats to version 1.33
5407 25. Include certificate and key file names in error message when GnuTLS fails
5408 to set them up, because the GnuTLS error message doesn't include the name
5409 of the failing file when there is a problem reading it.
5411 26. Expand error message when OpenSSL has problems setting up cert/key files.
5414 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5415 (this can affect the format of dates).
5417 28. exim_tidydb, when checking for the continued existence of a message for
5418 which it has found a message-specific retry record, was not finding
5419 messages that were in split spool directories. Consequently, it was
5420 deleting retry records that should have stayed in existence.
5422 29. eximstats updated to version 1.35
5423 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5424 1.35 - bugfix such that pie charts by volume are generated correctly
5426 30. The SPA authentication driver was not abandoning authentication and moving
5427 on to the next authenticator when an expansion was forced to fail,
5428 contradicting the general specification for all authenticators. Instead it
5429 was generating a temporary error. It now behaves as specified.
5431 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5432 (the order specifies the preference for clients). The order is now AES256,
5433 AES128, 3DES, ARCFOUR128.
5435 31. Small patch to Sieve code - explicitly set From: when generating an
5438 32. Exim crashed if a remote delivery caused a very long error message to be
5439 recorded - for instance if somebody sent an entire SpamAssassin report back
5440 as a large number of 550 error lines. This bug was coincidentally fixed by
5441 increasing the size of one of Exim's internal buffers (big_buffer) that
5442 happened as part of the Exiscan merge. However, to be on the safe side, I
5443 have made the code more robust (and fixed the comments that describe what
5446 33. Some experimental protocols are using DNS PTR records for new purposes. The
5447 keys for these records are domain names, not reversed IP addresses. The
5448 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5449 leaves it alone. Component reversal etc. now happens only for IP addresses.
5452 34. The host_aton() function is supposed to be passed a string that is known
5453 to be a valid IP address. However, in the case of IPv6 addresses, it was
5454 not checking this. This is a hostage to fortune. Exim now panics and dies
5455 if the condition is not met. A case was found where this could be provoked
5456 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5457 components; fortuitously, this particular loophole had already been fixed
5458 by change 4.50/55 or 4.44/33 above.
5460 If there are any other similar loopholes, the new check in host_aton()
5461 itself should stop them being exploited. The report I received stated that
5462 data on the command line could provoke the exploit when Exim was running as
5463 exim, but did not say which command line option was involved. All I could
5464 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5465 running as the user.
5468 35. There was a buffer overflow vulnerability in the SPA authentication code
5469 (which came originally from the Samba project). I have added a test to the
5470 spa_base64_to_bits() function which I hope fixes it.
5473 36. The daemon start-up calls getloadavg() while still root for those OS that
5474 need the first call to be done as root, but it missed one case: when
5475 deliver_queue_load_max is set with deliver_drop_privilege. This is
5476 necessary for the benefit of the queue runner, because there is no re-exec
5477 when deliver_drop_privilege is set.
5479 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5480 was in use, was not putting the data itself into the right store pool;
5481 consequently, it could be overwritten for a subsequent message in the same
5482 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5485 38. Sometimes the final signoff response after QUIT could fail to get
5486 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5487 < 0 before doing a fflush(). This bug looks as though it goes back to the
5488 introduction of TLS in release 3.20, but "sometimes" must have been rare
5489 because the tests only now provoked it.
5495 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5496 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5497 Now it returns an integer. A number of calls were still expecting a T/F
5498 return. Fortuitously, in all cases, the tests worked in OK situations,
5499 which is the norm. However, things would have gone wrong on any write
5500 failures on the smtp file descriptor. This function is used when sending
5501 messages over SMTP and also when doing verify callouts.
5503 2. When Exim is called to do synchronous delivery of a locally submitted
5504 message (the -odf or -odi options), it no longer closes stderr before doing
5507 3. Implemented the mua_wrapper option.
5509 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5511 5. Implemented the functions header_remove(), header_testname(),
5512 header_add_at_position(), and receive_remove_recipient(), and exported them
5515 6. If an ACL "warn" statement specified the addition of headers, Exim already
5516 inserted X-ACL-Warn: at the start if there was no header name. However, it
5517 was not making this test for the second and subsequent header lines if
5518 there were newlines in the string. This meant that an invalid header could
5519 be inserted if Exim was badly configured.
5521 7. Allow an ACL "warn" statement to add header lines at the start or after all
5522 the Received: headers, as well as at the end.
5524 8. Added the rcpt_4xx retry error code.
5526 9. Added postmaster_mailfrom=xxx to callout verification option.
5528 10. Added mailfrom=xxxx to the callout verification option, for verify=
5531 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5532 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5534 12. Inserted some casts to stop certain compilers warning when using pointer
5535 differences as field lengths or precisions in printf-type calls (mostly
5536 affecting debugging statements).
5538 13. Added optional readline() support for -be (dynamically loaded).
5540 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5541 same clock tick as a message's arrival, so that its received time was the
5542 same as the "first fail" time on the retry record, and that message
5543 remained on the queue past the ultimate address timeout, every queue runner
5544 would try a delivery (because it was past the ultimate address timeout) but
5545 after another failure, the ultimate address timeout, which should have then
5546 bounced the address, did not kick in. This was a "< instead of <=" error;
5547 in most cases the first failure would have been in the next clock tick
5548 after the received time, and all would be well.
5550 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5551 being recognized when the domain list was tested by the match_domain
5552 condition in an expansion string.
5554 16. Added the ${str2b64: operator.
5556 17. Exim was always calling setrlimit() to set a large limit for the number of
5557 processes, without checking whether the existing limit was already
5558 adequate. (It did check for the limit on file descriptors.) Furthermore,
5559 errors from getrlimit() and setrlimit() were being ignored. Now they are
5560 logged to the main and panic logs, but Exim does carry on, to try to do its
5561 job under whatever limits there are.
5563 18. Imported PCRE 5.0.
5565 19. Trivial typo in log message " temporarily refused connection" (the leading
5568 20. If the log selector return_path_on_delivery was set and an address was
5569 redirected to /dev/null, the delivery process crashed because it assumed
5570 that a return path would always be set for a "successful" delivery. In this
5571 case, the whole delivery is bypassed as an optimization, and therefore no
5574 21. Internal re-arrangement: the function for sending a challenge and reading
5575 a response while authentication was assuming a zero-terminated challenge
5576 string. It's now changed to take a pointer and a length, to allow for
5577 binary data in such strings.
5579 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5581 23. Exim was not respecting finduser_retries when seeking the login of the
5582 uid under which it was called; it was always trying 10 times. (The default
5583 setting of finduser_retries is zero.) Also, it was sleeping after the final
5584 failure, which is pointless.
5586 24. Implemented tls_on_connect_ports.
5588 25. Implemented acl_smtp_predata.
5590 26. If the domain in control=submission is set empty, Exim assumes that the
5591 authenticated id is a complete email address when it generates From: or
5592 Sender: header lines.
5594 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5595 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5596 chown and chgrp in /bin and hostname in /usr/bin.
5598 28. Exim was keeping the "process log" file open after each use, just as it
5599 does for the main log. This opens the possibility of it remaining open for
5600 long periods when the USR1 signal hits a daemon. Occasional processlog
5601 errors were reported, that could have been caused by this. Anyway, it seems
5602 much more sensible not to leave this file open at all, so that is what now
5605 29. The long-running daemon process does not normally write to the log once it
5606 has entered its main loop, and it closes the log before doing so. This is
5607 so that log files can straightforwardly be renamed and moved. However,
5608 there are a couple of unusual error situations where the daemon does write
5609 log entries, and I had neglected to close the log afterwards.
5611 30. The text of an SMTP error response that was received during a remote
5612 delivery was being truncated at 512 bytes. This is too short for some of
5613 the long messages that one sometimes sees. I've increased the limit to
5616 31. It is now possible to make retry rules that apply only when a message has a
5617 specific sender, in particular, an empty sender.
5619 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5620 it possible to be selective about when SMTP synchronization is enforced.
5622 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5624 32. Implemented hosts_connection_nolog.
5626 33. Added an ACL for QUIT.
5628 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5631 35. Added mailbox_size and mailbox_filecount to appendfile.
5633 36. Added control = no_multiline_responses to ACLs.
5635 37. There was a bug in the logic of the code that waits for the clock to tick
5636 in the case where the clock went backwards by a substantial amount such
5637 that the microsecond fraction of "now" was more than the microsecond
5638 fraction of "then" (but the whole seconds number was less).
5640 38. Added support for the libradius Radius client library this is found on
5641 FreeBSD (previously only the radiusclient library was supported).
5647 1. When certain lookups returned multiple values in the form name=value, the
5648 quoting of the values was not always being done properly. Specifically:
5649 (a) If the value started with a double quote, but contained no whitespace,
5651 (b) If the value contained whitespace other than a space character (i.e.
5652 tabs or newlines or carriage returns) it was not quoted.
5653 This fix has been applied to the mysql and pgsql lookups by writing a
5654 separate quoting function and calling it from the lookup code. The fix
5655 should probably also be applied to nisplus, ibase and oracle lookups, but
5656 since I cannot test any of those, I have not disturbed their existing code.
5658 2. A hit in the callout cache for a specific address caused a log line with no
5659 reason for rejecting RCPT. Now it says "Previous (cached) callout
5660 verification failure".
5662 3. There was an off-by-one bug in the queryprogram router. An over-long
5663 return line was truncated at 256 instead of 255 characters, thereby
5664 overflowing its buffer with the terminating zero. As well as fixing this, I
5665 have increased the buffer size to 1024 (and made a note to document this).
5667 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5668 when Exim is waiting for an SMTP response from a remote server, Exim
5669 restarts its select() call on the socket, thereby resetting its timeout.
5670 This is not a problem when such interrupts are rare. Somebody set up a cron
5671 job to run exiwhat every 2 minutes, which is less than the normal select()
5672 timeout (5 or 10 minutes). This meant that the select() timeout never
5673 kicked in because it was always reset. I have fixed this by comparing the
5674 time when an interrupt arrives with the time at the start of the first call
5675 to select(). If more time than the timeout has elapsed, the interrupt is
5676 treated as a timeout.
5678 5. Some internal re-factoring in preparation for the addition of Sieve
5679 extensions (by MH). In particular, the "personal" test is moved to a
5680 separate function, and given an option for scanning Cc: and Bcc: (which is
5681 not set for Exim filters).
5683 6. When Exim created an email address using the login of the caller as the
5684 local part (e.g. when creating a From: or Sender: header line), it was not
5685 quoting the local part when it contained special characters such as @.
5687 7. Installed new OpenBSD configuration files.
5689 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5690 try to make them clearer.
5692 9. Callout options, other than the timeout value, were being ignored when
5693 verifying sender addresses in header lines. For example, when using
5695 verify = header_sender/callout=no_cache
5697 the cache was (incorrectly) being used.
5699 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5700 only the Cygwin environment.
5702 11. When return_path_on_delivery was set as a log selector, if different remote
5703 addresses in the same message used different return paths and parallel
5704 remote delivery occurred, the wrong values would sometimes be logged.
5705 (Whenever a remote delivery process finished, the return path value from
5706 the most recently started remote delivery process was logged.)
5708 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5709 lines when AUTH and/or TLS are in use. This is the "received protocol"
5710 field. Exim used to use "asmtp" for authenticated SMTP, without any
5711 indication (in the protocol name) for TLS use. Now it follows the RFC and
5712 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5713 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5714 names appear in log lines as well as in Received: header lines.
5716 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5717 extensions, and comparison tests, and to fix some bugs.
5719 14. Changes to the "personal" filter test:
5721 (1) The test was buggy in that it was just doing the equivalent of
5722 "contains" tests on header lines. For example, if a user's address was
5723 anne@some.where, the "personal" test would incorrectly be true for
5725 To: susanne@some.where
5727 This test is now done by extracting each address from the header in turn,
5728 and checking the entire address. Other tests that are part of "personal"
5729 are now done using regular expressions (for example, to check local parts
5730 of addresses in From: header lines).
5732 (2) The list of non-personal local parts in From: addresses has been
5733 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5734 taken from the Sieve specification recommendations.
5736 (3) If the message contains any header line starting with "List-" it is
5737 treated as non-personal.
5739 (4) The test for "circular" in the Subject: header line has been removed
5740 because it now seems ill-conceived.
5742 15. Minor typos in src/EDITME comments corrected.
5744 16. Installed latest exipick from John Jetmore.
5746 17. If headers_add on a router specified a text string that was too long for
5747 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5748 of string_sprintf() is now avoided.
5750 18. $message_body_size was not set (it was always zero) when running the DATA
5751 ACL and the local_scan() function.
5753 19. For the "mail" command in an Exim filter, no default was being set for
5754 the once_repeat time, causing a random time value to be used if "once" was
5755 specified. (If the value happened to be <= 0, no repeat happened.) The
5756 default is now 0s, meaning "never repeat". The "vacation" command was OK
5757 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5758 (I found it when inspecting the code).
5760 20. There is now an overall timeout for performing a callout verification. It
5761 defaults to 4 times the callout timeout, which applies to individual SMTP
5762 commands during the callout. The overall timeout applies when there is more
5763 than one host that can be tried. The timeout is checked before trying the
5764 next host. This prevents very long delays if there are a large number of
5765 hosts and all are timing out (e.g. when the network connections are timing
5766 out). The value of the overall timeout can be changed by specifying an
5767 additional sub-option for "callout", called "maxwait". For example:
5769 verify = sender/callout=5s,maxwait=20s
5771 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5772 to the end before writing, but this should make it even safer).
5774 22. Exim was forgetting that it had advertised PIPELINING for the second and
5775 subsequent messages on an SMTP connection. It was also not resetting its
5776 memory on STARTTLS and an internal HELO.
5778 23. When Exim logs an SMTP synchronization error within a session, it now
5779 records whether PIPELINING has been advertised or not.
5781 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5782 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5783 rather than long int.
5785 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5787 26. Added the never_mail option to autoreply.
5793 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5794 crash if the getsockname() call failed; this can happen if a connection is
5795 closed very soon after it is established. The problem was simply in the
5796 order in which certain operations were done, causing Exim to try to write
5797 to the SMTP stream before it had set up the file descriptor. The bug has
5798 been fixed by making things happen in the correct order.
5804 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5805 before the connection was closed, thus losing the rejection response.
5807 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5808 some early Solaris releases, but causes trouble in current releases where
5809 socklen_t is defined.
5811 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5814 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5817 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5818 that releasing the top of it at the end releases what was used for sub-
5819 expansions (unless the block got too big). However, discard this block if
5820 the first thing is a variable or header, so that we can use its block when
5821 it is dynamic (useful for very large $message_headers, for example).
5823 6. Lookups now cache *every* query, not just the most recent. A new, separate
5824 store pool is used for this. It can be recovered when all lookup caches are
5825 flushed. Lookups now release memory at the end of their result strings.
5826 This has involved some general refactoring of the lookup sources.
5828 7. Some code has been added to the store_xxx() functions to reduce the amount
5829 of flapping under certain conditions.
5831 8. log_incoming_interface used to affect only the <= reception log lines. Now
5832 it causes the local interface and port to be added to several more SMTP log
5833 lines, for example "SMTP connection from", and rejection lines.
5835 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5837 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5839 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5840 could be overwritten at the end of the current message (or the start of a
5841 new message if it was set in a HELO ACL). The value is now preserved for
5842 the duration of the SMTP connection.
5844 12. If a transport had a headers_rewrite setting, and a matching header line
5845 contained an unqualified address, that address was qualified, even if it
5846 did not match any rewriting rules. The underlying bug was that the values
5847 of the flags that permit the existence of unqualified sender and recipient
5848 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5849 non-local messages, and by -bnq for local messages) were not being
5850 preserved with the message after it was received.
5852 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5853 "next input=" as part of the text comprising the host identity instead of
5854 the correct text. The code was using the same buffer for two different
5855 strings. However, depending on which order the printing function evaluated
5856 its arguments, the bug did not always show up. Under Linux, for example, my
5857 test suite worked just fine.
5859 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5860 doesn't work with some older versions of Perl. It has been changed to "my",
5861 which in any case is probably the better facility to use.
5863 15. A really picky compiler found some instances of statements for creating
5864 error messages that either had too many or two few arguments for the format
5867 16. The size of the buffer for calls to the DNS resolver has been increased
5868 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5869 for addresses that have a lot of PTR records. This alleviates a problem; it
5870 does not fully solve it.
5872 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5873 buffer now truncates the list and logs the incident, which is the same
5874 action as happens when Exim is looking up a host name and its aliases.
5875 Previously in this situation something unpredictable would happen;
5876 sometimes it was "internal error: store_reset failed".
5878 18. If a server dropped the connection unexpectedly when an Exim client was
5879 using GnuTLS and trying to read a response, the client delivery process
5880 crashed while trying to generate an error log message.
5882 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5883 string, for example:
5885 warn message = H1: something\nH2: something
5887 the text was added as a single header line from Exim's point of view
5888 though it ended up OK in the delivered message. However, searching for the
5889 second and subsequent header lines using $h_h2: did not work. This has been
5890 fixed. Similarly, if a system filter added multiple headers in this way,
5891 the routers could not see them.
5893 20. Expanded the error message when iplsearch is called with an invalid key to
5894 suggest using net-iplsearch in a host list.
5896 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5897 ACLs are no longer actually imposed (and a message to that effect is
5900 22. If a "gecos" field in a passwd entry contained escaped characters, in
5901 particular, if it contained a \" sequence, Exim got it wrong when building
5902 a From: or a Sender: header from that name. A second bug also caused
5903 incorrect handling when an unquoted " was present following a character
5904 that needed quoting.
5906 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5907 was not being matched caselessly.
5909 24. Arranged for all hyphens in the exim.8 source to be escaped with
5912 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5913 properly. Recipient callout cache records were still being keyed to include
5914 the sender, even when use_sender was set false. This led to far more
5915 callouts that were necessary. The sender is no longer included in the key
5916 when use_sender is false.
5918 26. Added "control = submission" modifier to ACLs.
5920 27. Added the ${base62d: operator to decode base 62 numbers.
5922 28. dnsdb lookups can now access SRV records.
5924 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5925 the configuration file.
5927 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5928 (-v) mode. This makes the output for a verbose queue run more intelligible.
5930 31. Added a use_postmaster feature to recipient callouts.
5932 32. Added the $body_zerocount variable, containing the number of binary zero
5933 bytes in the message body.
5935 33. The time of last modification of the "new" subdirectory is now used as the
5936 "mailbox time last read" when there is a quota error for a maildir
5939 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5941 35. Added +ignore_unknown as a special item in host lists.
5943 36. Code for decoding IPv6 addresses in host lists is now included, even if
5944 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5945 address was recognized as an IP address, but was then not correctly decoded
5946 into binary, causing unexpected and incorrect effects when compared with
5953 1. Very minor rewording of debugging text in manualroute to say "list of
5954 hosts" instead of "hostlist".
5956 2. If verify=header_syntax was set, and a header line with an unqualified
5957 address (no domain) and a large number of spaces between the end of the
5958 name and the colon was received, the reception process suffered a buffer
5959 overflow, and (when I tested it) crashed. This was caused by some obsolete
5960 code that should have been removed. The fix is to remove it!
5962 3. When running in the test harness, delay a bit after writing a bounce
5963 message to get a bit more predictability in the log output.
5965 4. Added a call to search_tidyup() just before forking a reception process. In
5966 theory, someone could use a lookup in the expansion of smtp_accept_max_
5967 per_host which, without the tidyup, could leave open a database connection.
5969 5. Added the variables $recipient_data and $sender_data which get set from a
5970 lookup success in an ACL "recipients" or "senders" condition, or a router
5971 "senders" option, similar to $domain_data and $local_part_data.
5973 6. Moved the writing of debug_print from before to after the "senders" test
5976 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5977 problems for message scanning, either using a data ACL, or using
5978 local_scan() because the Received: header was not generated till after they
5979 were called (in order to set the time as the time of reception completion).
5980 I have revised the way this works. The header is now generated after the
5981 body is received, but before the ACL or local_scan() are called. After they
5982 are run, the timestamp in the header is updated.
5988 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5989 before starting a queue runner without re-exec. This happened only when
5990 deliver_drop_privilege was set or when the Exim user was set to root. The
5991 effect of the bug was that timeouts during subsequent deliveries caused
5992 crashes instead of being properly handled. The handler is now left at its
5993 default (and expected) setting.
5995 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5996 message, again when deliver_drop_privilege is set or Exim is run as root.
5997 The bug described in (1) was not present in this case, but the tidying up
5998 of the other signals was missing. I have made the two cases consistent.
6000 3. The ignore_target_hosts setting on a manualroute router was being ignored
6001 for hosts that were looked up using the /MX notation.
6003 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6006 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6007 operated on the sender address. After changing the $sender_address to <>
6008 for the sender address verify, Exim was re-instated it as the original
6009 (before rewriting) address, but remembering that it had rewritten it, so it
6010 wasn't rewriting it again. This bug also had the effect of breaking the
6011 sender address verification caching when the sender address was rewritten.
6013 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6014 This has been changed so that if the ip literal address matches
6015 ignore_target_hosts, the router declines.
6017 7. Added expansion conditions match_domain, match_address, and match_local_
6018 part (NOT match_host).
6020 8. The placeholder for the Received: header didn't have a length field set.
6022 9. Added code to Exim itself and to exim_lock to test for a specific race
6023 condition that could lead to file corruption when using MBX delivery. The
6024 issue is with the lockfile that is created in /tmp. If this file is removed
6025 after a process has opened it but before that process has acquired a lock,
6026 there is the potential for a second process to recreate the file and also
6027 acquire a lock. This could lead to two Exim processes writing to the file
6028 at the same time. The added code performs the same test as UW imapd; it
6029 checks after acquiring the lock that its file descriptor still refers to
6030 the same named file.
6032 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6033 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6036 11. Added the smtp_active_hostname option. If used, this will typically be made
6037 to depend on the incoming interface address. Because $interface_address is
6038 not set up until the daemon has forked a reception process, error responses
6039 that can happen earlier (such as "too many connections") no longer contain
6042 12. If an expansion in a condition on a "warn" statement fails because a lookup
6043 defers, the "warn" statement is abandoned, and the next ACL statement is
6044 processed. Previously this caused the whole ACL to be aborted.
6046 13. Added the iplsearch lookup type.
6048 14. Added ident_timeout as a log selector.
6050 15. Added tls_certificate_verified as a log selector.
6052 16. Added a global option tls_require_ciphers (compare the smtp transport
6053 option of the same name). This controls incoming TLS connections.
6055 17. I finally figured out how to make tls_require_ciphers do a similar thing
6056 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6057 before starting the TLS session.
6059 18. Tabs are now shown as \t in -bP output.
6061 19. If the log selector return_path_on_delivery was set, Exim crashed when
6062 bouncing a message because it had too many Received: header lines.
6064 20. If two routers both had headers_remove settings, and the first one included
6065 a superfluous trailing colon, the final name in the first list and the
6066 first name in the second list were incorrectly joined into one item (with a
6067 colon in the middle).
6073 1. Added -C and -D options to the exinext utility, mainly to make it easier
6074 to include in the automated testing, but these could be helpful when
6075 multiple configurations are in use.
6077 2. The exinext utility was not formatting the output nicely when there was
6078 an alternate port involved in the retry record key, nor when there was a
6079 message id as well (for retries that were specific to a specific message
6080 and a specific host). It was also confused by IPv6 addresses, because of
6081 the additional colons they contain. I have fixed the IPv4 problem, and
6082 patched it up to do a reasonable job for IPv6.
6084 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6085 delivery, the log line now contains "pipelined" if PIPELINING was used.
6087 4. An SMTP transport process used to panic and die if the bind() call to set
6088 an explicit outgoing interface failed. This has been changed; it is now
6089 treated in the same way as a connect() failure.
6091 5. A reference to $sender_host_name in the part of a conditional expansion
6092 that was being skipped was still causing a DNS lookup. This no longer
6095 6. The def: expansion condition was not recognizing references to header lines
6096 that used bh_ and bheader_.
6098 7. Added the _cache feature to named lists.
6100 8. The code for checking quota_filecount in the appendfile transport was
6101 allowing one more file than it should have been.
6103 9. For compatibility with Sendmail, the command line option
6111 and sets the incoming protocol and host name (for trusted callers). The
6112 host name and its colon can be omitted when only the protocol is to be set.
6113 Note the Exim already has two private options, -pd and -ps, that refer to
6114 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6115 "s", but I don't think that's a major issue.
6117 10. A number of refactoring changes to the code, none of which should affect
6120 (a) The number of logging options was getting close to filling up the
6121 32-bit word that was used as a bit map. I have split them into two classes:
6122 those that are passed in the argument to log_write(), and those that are
6123 only ever tested independently outside of that function. These are now in
6124 separate 32-bit words, so there is plenty of room for expansion again.
6125 There is no change in the user interface or the logging behaviour.
6127 (b) When building, for example, log lines, the code previously used a
6128 macro that called string_cat() twice, in order to add two strings. This is
6129 not really sufficiently general. Furthermore, there was one instance where
6130 it was actually wrong because one of the argument was used twice, and in
6131 one call a function was used. (As it happened, calling the function twice
6132 did not affect the overall behaviour.) The macro has been replaced by a
6133 function that can join an arbitrary number of extra strings onto a growing
6136 (c) The code for expansion conditions now uses a table and a binary chop
6137 instead of a serial search (which was left over from when there were very
6138 few conditions). Also, it now recognizes conditions like "pam" even when
6139 the relevant support is not compiled in: a suitably worded error message is
6140 given if an attempt is made to use such a condition.
6142 11. Added ${time_interval:xxxxx}.
6144 12. A bug was causing one of the ddress fields not to be passed back correctly
6145 from remote delivery subprocesses. The field in question was not being
6146 subsequently used, so this caused to problems in practice.
6148 13. Added new log selectors queue_time and deliver_time.
6150 14. Might have fixed a bug in maildirsizefile handling that threw up
6151 "unexpected character" debug warnings, and recalculated the data
6152 unnecessarily. In any case, I expanded the warning message to give more
6155 15. Added the message "Restricted characters in address" to the statements in
6156 the default ACL that block characters like @ and % in local parts.
6158 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6159 Three changes have been made:
6161 (a) There was a serious bug; a negative response to MAIL caused the whole
6162 recipient domain to be cached as invalid, thereby blocking all messages
6163 to all local parts at the same domain, from all senders. This bug has
6164 been fixed. The domain is no longer cached after a negative response to
6165 MAIL if the sender used is not empty.
6167 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6170 (c) A new callout option, "use_sender" has been added for people who want
6171 the modified behaviour.
6177 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6180 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6181 indeed breaks things for older releases.
6183 3. Added additional logging to the case where there is a problem reading data
6184 from a filter that is running in a subprocess using a pipe, in order to
6185 try to track down a specific problem.
6187 4. Testing facility fudge: when running in the test harness and attempting
6188 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6189 getting "No route to host". Convert this to a timeout.
6191 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6194 6. Some OS don't have socklen_t but use size_t instead. This affects the
6195 fifth argument of getsockopt() amongst other things. This is now
6196 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6197 can be set for individual OS. I have set it for SunOS5, OSF1, and
6198 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6199 some earlier ones do not.
6201 7. Change 4.30/15 was not doing the test caselessly.
6203 8. The standard form for an IPv6 address literal was being rejected by address
6204 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6205 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6206 this, as well as the form without the "IPv6" on the front (but only when
6207 address literals are enabled, of course).
6209 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6211 10. Exim crashed if a message with an empty sender address specified by -f
6212 encountered a router with an errors_to setting. This could be provoked only
6213 by a command such as
6217 where an empty string was supplied; "<>" did not hit this bug.
6219 11. Installed PCRE release 4.5.
6221 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6222 remained set. It is now erased.
6224 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6225 times from message ids (which are base 36 rather than the normal 62).
6227 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6228 were being counted as actual protocol errors, and logged if the log
6229 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6230 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6231 and DATA following a set of rejected RCPTs do not count as protocol errors.
6232 In other words, Exim assumes they were pipelined, though this may not
6233 actually be the case. Of course, in all cases the client gets an
6234 appropriate error code.
6236 15. If a lookup fails in an ACL condition, a message about the failure may
6237 be available; it is used if testing the ACL cannot continue, because most
6238 such messages specify what the cause of the deferral is. However, some
6239 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6240 that caused an old message to be retained and used if a later statement
6241 caused a defer, replacing the real cause of the deferral.
6243 16. If an IP address had so many PTR records that the DNS lookup buffer
6244 was not large enough to hold them, Exim could crash while trying to process
6245 the truncated data. It now detects and logs this case.
6247 17. Further to 4.21/58, another change has been made: if (and only if) the
6248 first line of a message (the first header line) ends with CRLF, a bare LF
6249 in a subsequent header line has a space inserted after it, so as not to
6250 terminate the header.
6252 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6253 unnecessarily, used atoi() instead of strtol(), and didn't check the
6254 termination when getting file sizes from file names by regex.
6256 19. Completely re-implemented the support for maildirsize files, in the light
6257 of a number of problems with the previous contributed implementation
6258 (4.30/29). In particular:
6260 . If the quota is zero, the maildirsize file is maintained, but no quota is
6263 . If the maildir directory does not exist, it is created before any attempt
6264 to write a maildirsize file.
6266 . The quota value in the file is just a cache; if the quota is changed in
6267 the transport, the new value overrides.
6269 . A regular expression is available for excluding directories from the
6272 20. The autoreply transport checks the characters in options that define the
6273 message's headers; it allows continued headers, but it was checking with
6274 isspace() after an embedded newline instead of explicitly looking for a
6277 21. If all the "regular" hosts to which an address was routed had passed their
6278 expiry times, and had not reached their retry times, the address was
6279 bounced, even if fallback hosts were defined. Now Exim should go on to try
6282 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6283 equivalent code in the SMTP transport. Some hosts send humungous responses
6284 to HELO/EHLO, more than 1024 it seems.
6286 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6287 gives compiler warnings in some environments. I've now done it "properly",
6290 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6291 (because of problems with the built-in one) was declared to return uschar *
6292 instead of char *, causing compiler failure.
6294 25. Fixed a file descriptor leak when processing alias/forward files.
6296 26. Fixed a minor format string issue in dbfn.c.
6298 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6300 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6301 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6304 29. When a sender address is verified, it is cached, to save repeating the test
6305 when there is more than one recipient in a message. However, when the
6306 verification involves a callout, it is possible for different callout
6307 options to be set for different recipients. It is too complicated to keep
6308 track of this in the cache, so now Exim always runs a verification when a
6309 callout is required, relying on the callout cache for the optimization.
6310 The overhead is duplication of the address routing, but this should not be
6313 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6314 to be verified with callout=postmaster, and the main callout worked but the
6315 postmaster check failed, the verification correctly failed. However, if a
6316 subsequent RCPT command asked for sender verification *without* the
6317 postmaster check, incorrect caching caused this verification also to fail,
6320 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6321 it was not caching the DNS options (qualify_single, search_parents) that
6322 were used when the lookup failed. A subsequent lookup with different
6323 options therefore always gave the same answer, though there were cases
6324 where it should not have. (Example: a "domains = !$mx_any" option on a
6325 dnslookup router: the "domains" option is always processed without any
6326 widening, but the router might have qualify_single set.) Now Exim uses the
6327 cached value only when the same options are set.
6329 32. Added John Jetmore's "exipick" utility to the distribution.
6331 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6332 than a timeout (e.g. a certificate is required, and is not provided), an
6333 Exim server now closes the connection immediately. Previously it waited for
6334 the client to close - but if the client is SSL, it seems that they each
6335 wait for each other, leading to a delay before one of them times out.
6337 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6338 maintained 0.8.x compatibility because I don't think many are using it, and
6339 it is clearly obsolete.
6341 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6344 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6345 client certificate was expired. A simple patch fixes this, though I don't
6346 understand the full logic of why the verify callback is called multiple
6349 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6350 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6351 which causes problems with some clients (such as the Certicom SSL Plus
6352 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6353 disables the coutermeasure allowing Eudora to connect."
6355 38. Exim was not checking that a write() to a log file succeeded. This could
6356 lead to Bad Things if a log got too big, in particular if it hit a file
6357 size limit. Exim now panics and dies if it cannot write to a log file, just
6358 as it does if it cannot open a log file.
6360 39. Modified OS/Makefile-Linux so that it now contains
6362 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6364 The two -D definitions ensure that Exim is compiled with large file
6365 support, which makes it possible to handle log files that are bigger than
6368 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6369 instance) a domain was checked against a named list that involved a lookup,
6370 causing $domain_data to be set, then another domain was checked against the
6371 same list, then the first domain was re-checked, the value of $domain_data
6372 after the final check could be wrong. In particular, if the second check
6373 failed, it could be set empty. This bug probably also applied to
6376 41. The strip_trailing_dot option was not being applied to the address given
6377 with the -f command-line option.
6379 42. The code for reading a message's header from the spool was incrementing
6380 $received_count, but never initializing it. This meant that the value was
6381 incorrect (doubled) while delivering a message in the same process in which
6382 it was received. In the most common configuration of Exim, this never
6383 happens - a fresh exec is done - but it can happen when
6384 deliver_drop_privilege is set.
6386 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6387 it now includes up to 150 characters of the unexpected data in the log
6390 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6391 and building data strings. The size of both of these buffers was 10 000
6392 bytes - far larger than anybody would *ever* want, thought I. Needless to
6393 say, somebody hit the limit. I have increased the maximum line length to
6394 20 000 and the maximum data length of concatenated lines to 100 000. I have
6395 also fixed two bugs, because there was no checking on these buffers. Tsk,
6396 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6397 buffer is too small.
6399 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6400 lsearch lookups. Now it does.
6402 46. When parsing a route_list item in a manualroute router, a fixed-length
6403 buffer was used for the list of hosts. I made this 1024 bytes long,
6404 thinking that nobody would ever have a list of hosts that long. Wrong.
6405 Somebody had a whole pile of complicated expansion conditions, and the
6406 string was silently truncated, leading to an expansion error. It turns out
6407 that it is easier to change to an unlimited length (owing to other changes
6408 that have happened since this code was originally written) than to build
6409 structure for giving a limitation error. The length of the item that
6410 expands into the list of hosts is now unlimited.
6412 47. The lsearch lookup could not handle data where the length of text line was
6413 more than 4095 characters. Such lines were truncated, leading to shortened
6414 data being returned. It should now handle lines of any length.
6416 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6417 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6420 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6421 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6422 no longer happens in comment lines. A list of replacements is now placed
6423 at the head of all of the source files, except those whose only change is
6424 to replace PERL_COMMAND in the very first #! line.
6426 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6427 messages on the queue, with a bottom-up merge sort, using code contributed
6428 by Michael Haardt. This should make operations like -bp somewhat faster on
6429 large queues. It won't affect queue runners, except when queue_run_in_order
6432 51. Installed eximstats 1.31 in the distribution.
6434 52. Added support for SRV lookups to the dnslookup router.
6436 53. If an ACL referred to $message_body or $message_body_end, the value was not
6437 reset for any messages that followed in the same SMTP session.
6439 54. The store-handling optimization for building very long strings was not
6440 differentiating between the different store pools. I don't think this
6441 actually made any difference in practice, but I've tidied it.
6443 55. While running the routers to verify a sender address, $sender_address
6444 was still set to the sender address. This is wrong, because when routing to
6445 send a bounce to the sender, it would be empty. Therefore, I have changed
6446 it so that, while verifying a sender address, $sender_address is set to <>.
6447 (There is no change to what happens when verifying a recipient address.)
6449 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6450 target A or AAAA records (if not already returned) without resetting the
6451 qualify_single or search_parents options of the DNS resolver. These are
6452 inappropriate in this case because the targets of MX and SRV records must
6453 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6454 target that, when qualified, matched something in the local domain. These
6455 two options are now turned off when doing these lookups.
6457 57. It seems that at least some releases of Reiserfs (which does not have the
6458 concept of a fixed number of inodes) returns zero and not -1 for the
6459 number of available inodes. This interacted badly with check_spool_inodes,
6460 which assumed that -1 was the "no such thing" setting. What I have done is
6461 to check that the total number of inodes is greater than zero before doing
6462 the test of how many are available.
6464 58. When a "warn" ACL statement has a log_message modifier, the message is
6465 remembered, and not repeated. This is to avoid a lot of repetition when a
6466 message has many recipients that cause the same warning to be written.
6467 However, Exim was preserving the list of already written lines for an
6468 entire SMTP session, which doesn't seem right. The memory is now reset if a
6469 new message is started.
6471 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6472 debugging output unless log_rewrite was also set.
6474 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6475 of (int)(handle) when we know that handle contains (void *)(-1).
6477 61. The Exim daemon panic-logs an error return when it closes the incoming
6478 connection. However "connection reset by peer" seems to be common, and
6479 isn't really an error worthy of noting specially, so that particular error
6482 62. When Exim is trying to find all the local interfaces, it used to panic and
6483 die if the ioctl to get the interface flags failed. However, it seems that
6484 on at least one OS (Solaris 9) it is possible to have an interface that is
6485 included in the list of interfaces, but for which you get a failure error
6486 for this call. This happens when the interface is not "plumbed" into a
6487 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6488 failure of the "get flags" call assumes that the interface is down.
6490 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6491 makes life easier for people who are doing arithmetic on fields extracted
6492 from dates, where you often get leading zeros that should not be
6493 interpreted as octal.
6495 64. Added qualify_domain to the redirect router, to override the global
6498 65. If a pathologically long header line contained very many addresses (the
6499 report of this problem mentioned 10 000) and each of them was rewritten,
6500 Exim could use up a very large amount of memory. (It kept on making new
6501 copies of the header line as it rewrote, and never released the old ones.)
6502 At the expense of a bit more processing, the header rewriting function has
6503 been changed so that it no longer eats memory in this way.
6505 66. The generation of the Received: header has been moved from the time that a
6506 message starts to be received, to the time that it finishes. The timestamp
6507 in the Received: header should now be very close to that of the <= log
6508 line. There are two side-effects of this change:
6510 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6511 logged header lines no longer include the local Received: line, because
6512 it has not yet been created. The same applies to a copy of the message
6513 that is returned to a non-SMTP sender when a message is rejected.
6515 (b) When a filter file is tested using -bf, no additional Received: header
6516 is added to the test message. After some thought, I decided that this
6519 This change does not affect the value of $received_for. It is still set
6520 after address rewriting, but before local_scan() is called.
6522 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6524 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6525 gave an unhelpful panic error message, and a defer error. I have managed to
6526 change this behaviour so that it now rejects any supplied certificate,
6527 which seems right, as the list of acceptable certificates is empty.
6529 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6530 gave an unhelpful defer error. I have not managed to make this reject any
6531 supplied certificates, but the error message it gives is "no certificate
6532 supplied", which is not helpful.
6534 70. exigrep's output now also includes lines that are not associated with any
6535 message, but which match the given pattern. Implemented by a patch from
6536 Martin Sluka, which also tidied up the Perl a bit.
6538 71. Recipient callout verification, like sender verification, was using <> in
6539 the MAIL FROM command. This isn't really the right thing, since the actual
6540 sender may affect whether the remote host accepts the recipient or not. I
6541 have changed it to use the actual sender in the callout; this means that
6542 the cache record is now keyed on a recipient/sender pair, not just the
6543 recipient address. There doesn't seem to be a real danger of callout loops,
6544 since a callout by the remote host to check the sender would use <>.
6545 [SEE ABOVE: changed after hitting problems.]
6547 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6548 temporary errors. However, in the case of such a code being given after
6549 the end of a data transmission (i.e. after ".") Exim was failing to write
6550 a retry record for the message. (Yes, there was some broken host that was
6551 actually sending 8xx at this point.)
6553 73. An unknown lookup type in a host list could cause Exim to panic-die when
6554 the list was checked. (An example that provoked this was putting <; in the
6555 middle of a list instead of at the start.) If this happened during a DATA
6556 ACL check, a -D file could be left lying around. This kind of configuration
6557 error no longer causes Exim to die; instead it causes a defer error. The
6558 incident is still logged to the main and panic logs.
6560 74. Buglet left over from Exim 3 conversion. The message "too many messages
6561 in one connection" was written to the rejectlog but not the mainlog, except
6562 when address rewriting (yes!) was being logged.
6564 75. Added write_rejectlog option.
6566 76. When a system filter was run not as root (that is, when system_filter_user
6567 was set), the values of the $n variables were not being returned to the
6568 main process; thus, they were not subsequently available in the $sn
6571 77. Added +return_path_on_delivery log selector.
6573 78. A connection timeout was being treated differently from recipients deferred
6574 when testing hosts_max_try with a message that was older than the host's
6575 retry timeout. (The host should not be counted, thus allowing all hosts to
6576 be tried at least once before bouncing.) This may have been the cause of an
6577 occasionally reported bug whereby a message would remain on the queue
6578 longer than the retry timeout, but would be bounced if a delivery was
6579 forced. I say "may" because I never totally pinned down the problem;
6580 setting up timeout/retry tests is difficult. See also the next item.
6582 79. The ultimate address timeout was not being applied to errors that involved
6583 a combination of host plus message (for example, a timeout on a MAIL
6584 command). When an address resolved to a number of possible hosts, and they
6585 were not all tried for each delivery (e.g. because of hosts_max_try), a
6586 message could remain on the queue longer than the retry timeout.
6588 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6591 81. Fixed an obscure SMTP outgoing bug which required at least the following
6592 conditions: (a) there was another message waiting for the same server;
6593 (b) the server returned 5xx to all RCPT commands in the first message so
6594 that the message was not completed; (c) the server dropped the connection
6595 or gave a negative response to the RSET that Exim sends to abort the
6596 transaction. The observed case was a dropped connection after DATA that had
6597 been sent in pipelining mode. That is, the server had advertised PIPELINING
6598 but was not implementing it correctly. The effect of the bug was incorrect
6599 behaviour, such as trying another host, and this could lead to a crash.
6605 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6606 and daemon.c were passed as pointers to ints; they should have been
6607 pointers to socklen_t variables (which are typically unsigned ints).
6609 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6612 3. Fixed a really odd bug that affected only the testing scheme; patching a
6613 certain fixed string in the binary changed the value of another string that
6614 happened to be identical to the end of the original first string.
6616 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6617 name", it returns that address as the IP address. On some operating
6618 systems (e.g. Solaris), it also passes back the IP address string as the
6619 "host name". However, on others (e.g. Linux), it passes back an empty
6620 string. Exim wasn't checking for this, and was changing the host name to an
6621 empty string, assuming it had been canonicalized.
6623 5. Although rare, it is permitted to have more than one PTR record for a given
6624 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6625 all the names associated with an address, because they do in Solaris.
6626 However, it seems that they do not in Linux for data that comes from the
6627 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6628 I found this out when I moved to a new Linux workstation and tried to run
6629 the Exim test suite.
6631 To get round this problem I have changed the code so that it now does its
6632 own call to the DNS to look up PTR records when searching for a host name.
6633 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6634 addresses that are only in /etc/hosts are still found.
6636 This behaviour is, however, controlled by an option called host_lookup_
6637 order, which defaults to "bydns:byaddr". If people want to use the other
6638 order, or indeed, just use one or the other means of lookup, they can
6639 specify it in this variable.
6641 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6642 some operating systems, this comes back from gethostbyaddr() as an empty
6643 string, and this is what Exim used to test for. However, it seems that in
6644 other systems, "." is yielded. Exim now tests for this case too.
6646 7. The values of check_spool_space and check_log_space are now held internally
6647 as a number of kilobytes instead of an absolute number of bytes. If a
6648 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6649 kilobyte. This means that much larger values can be stored.
6651 8. Exim monitor: an attempt to get the action menu when not actually pointing
6652 at a message produces an empty menu entitled "No message selected". This
6653 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6654 no entries in it ("Shell widget menu has zero width and/or height"). So I
6655 have added a single, blank menu entry in this case.
6657 9. Added ${quote_local_part.
6659 10. MIME decoding is now applied to the contents of Subject: header lines when
6662 11. Now that a reference to $sender_host_address automatically causes a reverse
6663 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6664 host lookup before query-style lookups in lists that might use this
6665 variable. This has therefore been abolished, and the "net-" prefix is no
6666 longer necessary for query-style lookups.
6668 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6669 have been a typo for LFLAGS, so it has been changed.
6671 13. The install script calls Exim with "-C /dev/null" in order to find the
6672 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6673 to be output. However, since Exim outputs its version number before the
6674 error, it didn't break the script. It just looked ugly. I fixed this by
6675 always allowing "-C /dev/null" if the caller is root.
6677 14. Ignore overlarge ACL variable number when reading spool file - insurance
6678 against a later release with more variables having written the file.
6680 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6681 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6682 this, as well as the form without the "IPv6" on the front.
6684 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6685 OS/Makefile-Darwin file.
6687 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6688 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6690 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6691 variable. However, some OS use ints for inodes. I've added cast to long int
6692 to get rid of the compiler warning.
6694 19. I had forgotten to lock out "/../" in configuration file names when
6695 ALT_CONFIG_PREFIX was set.
6697 20. Routers used for verification do not need to specify transports. However,
6698 if such a router generated a host list, and callout was configured, Exim
6699 crashed, because it could not find a port number from the (non-existent)
6700 transport. It now assumes port 25 in this circumstance.
6702 21. Added the -t option to exigrep.
6704 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6705 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6706 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6707 them is set, LOOKUP_LSEARCH is forced.
6709 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6711 24. Added sender and host information to the "rejected by local_scan()" log
6712 line; previously there was no indication of these.
6714 25. Added .include_if_exists.
6716 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6717 Linux. It turns out that not all file systems support this. Apparently some
6718 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6719 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6720 not supported on the file descriptor, is now ignored when Exim is trying to
6721 sync a directory. This applies only to Linux.
6723 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6725 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6726 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6729 29. Added support for maildirsize files from supplied patch (modified a bit).
6731 30. The use of :fail: followed by an empty string could lead Exim to respond to
6732 sender verification failures with (e.g.):
6734 550 Verification failed for <xxx>
6735 550 Sender verify failed
6737 where the first response line was missing the '-' that indicates it is not
6738 the final line of the response.
6740 31. The loop for finding the name of the user that called Exim had a hardwired
6741 limit of 10; it now uses the value of finduser_retries, which is used for
6742 all other user lookups.
6744 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6747 33. Exim was neglecting to zero errno before one call of strtol() when
6748 expanding a string and expecting an integer value. On some systems this
6749 resulted in spurious "integer overflow" errors. Also, it was casting the
6750 result into an int without checking.
6752 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6753 did not work. The code looks as if it has *never* worked, though it appears
6754 to have been documented since at least release 1.62. I have made it work.
6756 35. The "timeout_DNS" error in retry rules, also documented since at least
6757 1.62, also never worked. As it isn't clear exactly what this means, and
6758 clearly it isn't a major issue, I have abolished the feature by treating it
6759 as "timeout", and writing a warning to the main and panic logs.
6761 36. The display of retry rules for -brt wasn't always showing the error code
6764 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6765 timeout_connect_A, timeout_connect_MX.
6767 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6768 to the empty sender.
6770 39. The daemon was not analysing the content of -oX till after it had closed
6771 stderr and disconnected from the controlling terminal. This meant that any
6772 syntax errors were only noted on the panic log, and the return code from
6773 the command was 0. By re-arranging the code a little, I've made the
6774 decoding happen first, so such errors now appear on stderr, and the return
6775 code is 1. However, the actual setting up of the sockets still happens in
6776 the disconnected process, so errors there are still only recorded on the
6779 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6780 connections (as happens on some IP stacks) was logged at start up time as
6781 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6782 it from "IPv6 and IPv4", which means that two separate sockets are being
6785 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6786 says whether AF_INET or AF_INET6 was passed as an argument.
6788 42. Exiwhat output was messed up when time zones were included in log
6791 43. Exiwhat now gives more information about the daemon's listening ports,
6792 and whether -tls-on-connect was used.
6794 44. The "port" option of the smtp transport is now expanded.
6796 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6797 silently ignored. Now an error message is written to the main and panic
6800 46. There's a new ACL modifier called "logwrite" which writes to a log file
6801 as soon as it is encountered.
6803 47. Added $local_user_uid and $local_user_gid at routing time.
6805 48. Exim crashed when trying to verify a sender address that was being
6808 49. Exim was recognizing only a space character after ".include". It now also
6809 recognizes a tab character.
6811 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6812 extracting the relevant information from the specification. The man page no
6813 longer contains scrambled data for the -d option, and I've added a section
6814 at the front about calling Exim under different names.
6816 51. Added "extra_headers" argument to the "mail" command in filter files.
6818 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6821 53. Installed eximstats 1.29.
6823 54. Added transport_filter_timeout as a generic transport option.
6825 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6826 Cc: header lines. This was required by RFC 822, but it not required by RFC
6829 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6830 incoming messages that did not have them. Now it does so only if the
6831 message originates locally, that is, if there is no associated remote host
6832 address. When Resent- header lines are present, this applies to the Resent-
6833 lines rather than the non-Resent- lines.
6835 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6836 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6838 58. Messages for configuration errors now include the name of the main
6839 configuration file - useful now that there may be more than one file in a
6840 list (.included file names were always shown).
6842 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6843 for those rare installations that do not start the daemon as root or run it
6844 setuid root. I've cut out the call to initgroups() if the daemon is not
6847 60. The Exim user and group can now be bound into the binary as text strings
6848 that are looked up at the start of Exim's processing.
6850 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6852 62. Added $mailstore_basename variable.
6854 63. Installed patch to sieve.c from Michael Haardt.
6856 64. When Exim failed to open the panic log after failing to open the main log,
6857 the original message it was trying to log was written to stderr and debug
6858 output, but if they were not available (the usual case in production), it
6859 was lost. Now it is written to syslog before the two lines that record the
6860 failures to open the logs.
6862 65. Users' Exim filters run in subprocesses under the user's uid. It is
6863 possible for a "deliver" command or an alias in a "personal" command to
6864 provoke an address rewrite. If logging of address rewriting is configured,
6865 this fails because the process is not running as root or exim. There may be
6866 a better way of dealing with this, but for the moment (because 4.30 needs
6867 to be released), I have disabled address rewrite logging when running a
6868 filter in a non-root, non-exim process.
6874 1. The buildconfig auxiliary program wasn't quoting the value set for
6875 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6876 not defined. This bug was masked in 4.22 by the effect that was fixed in
6879 2. Some messages that were rejected after a message id was allocated were
6880 shown as "incomplete" by exigrep. It no longer does this for messages that
6881 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6883 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6884 have allow_domain_literals set, the ID did not get logged in the <= line.
6885 Domain literals are now always recognized in Message-ID: header lines.
6887 4. The first argument for a ${extract expansion item is the key name or field
6888 number. Leading and trailing spaces in this item were not being ignored,
6889 causing some misleading effects.
6891 5. When deliver_drop_privilege was set, single queue runner processes started
6892 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6893 same command in the process it spins off) were not dropping privilege.
6895 6. When the daemon running as "exim" started a queue runner, it always
6896 re-executed Exim in the spun-off process. This is a waste of effort when
6897 deliver_drop_privilege is set. The new process now just calls the
6898 queue-runner function directly.
6904 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6907 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6908 ignored. Though the use of -oP was forcing the writing of a pid file, it
6909 was always written to the default place.
6911 3. If the message "no IP address found for host xxxx" is generated during
6912 incoming verification, it is now followed by identification of the incoming
6913 connection (so you can more easily find what provoked it).
6915 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6917 5. Added some features to "harden" Exim a bit more against certain attacks:
6919 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6920 be put in Local/Makefile. This is like the never_users runtime option,
6921 but it cannot be overridden. The default setting is "root".
6923 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6924 prefix string with which any file named in a -C command line option
6927 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6928 is retained for -C and -D only if the caller of Exim is root. Without
6929 it, the exim user may also use -C and -D and retain privilege.
6931 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6932 command line option is disabled.
6934 6. Macro names set by the -D option must start with an upper case letter, just
6935 like macro names defined in the configuration file.
6937 7. Added "dereference=" facility to LDAP.
6939 8. Two instances of the typo "uknown" in the source files are fixed.
6941 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6942 the Configure-Makefile script screwed up while processing it.
6944 10. Incorporated PCRE 4.4.
6946 11. The SMTP synchronization check was not operating right at the start of an
6947 SMTP session. For example, it could not catch a HELO sent before the client
6948 waited for the greeting. There is now a check for outstanding input at the
6949 point when the greeting is written. Because of the duplex, asynchronous
6950 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6951 way, but not yet received, when the check is performed.
6953 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6954 on TCP/IP sockets, because this apparently causes some broken clients to
6957 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6958 unchanged) from the Cygwin maintainer.
6960 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6961 was supported instead of testing for mbx. Effectively a typo.
6963 15. The spa authenticator server code was not checking that the input it
6964 received was valid base64.
6966 16. The debug output line for the "set" modifier in ACLs was not showing the
6967 name of the variable that was being set.
6969 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6971 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6972 lookup on its own. Something else was needed to trigger the lookup. For
6973 example, a match in host_lookup or the need for a host name in a host list.
6974 Now, if $sender_host_name is referenced and the host name has not yet been
6975 looked up, a lookup is performed. If the lookup fails, the variable remains
6976 empty, and $host_lookup_failed is set to "1".
6978 19. Added "eqi" as a case-independent comparison operator.
6980 20. The saslauthd authentication condition could segfault if neither service
6981 nor realm was specified.
6983 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6984 error message that was logged was misleading, and incoming SMTP
6985 connections were dropped. The message is now more accurate, and temporary
6986 errors are given to SMTP connections.
6988 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6989 (except RSET) with a 421 error, until QUIT is received. However, it was
6990 failing to send a response to QUIT.
6992 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6993 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6994 ACL is now run after the helo_try_verify_hosts code.
6996 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6997 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6998 case-independent, so other case variants are also recognized). Apparently
6999 some systems use these upper case variants.
7001 25. If more than two messages were waiting for the same host, and a transport
7002 filter was specified for the transport, Exim sent two messages over the
7003 same TCP/IP connection, and then failed with "socket operation on non-
7004 socket" when it tried to send the third.
7006 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7008 27. The extern definition of crypt16() in expand.c was not being excluded when
7009 the OS had its own crypt16() function.
7011 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7012 as a preferred synonym for return_size_limit, both as an option and as an
7015 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7017 30. Changed the default configuration ACL to relax the local part checking rule
7018 for addresses that are not in any local domains. For these addresses,
7019 slashes and pipe symbols are allowed within local parts, but the sequence
7020 /../ is explicitly forbidden.
7022 31. SPA server authentication was not clearing the challenge buffer before
7025 32. log_message in a "warn" ACL statement was writing to the reject log as
7026 well as to the main log, which contradicts the documentation and doesn't
7027 seem right (because no rejection is happening). So I have stopped it.
7029 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7030 However, I am unable to do any testing of this.
7032 34. Fixed an infelicity in the appendfile transport. When checking directories
7033 for a mailbox, to see if any needed to be created, it was accidentally
7034 using path names with one or more superfluous leading slashes; tracing
7035 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7037 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7038 contents are added to the log line that is written for every discarded
7039 recipient. (Previously a log_message setting was ignored.)
7041 36. The ${quote: operator now quotes the string if it is empty.
7043 37. The install script runs exim in order to find its version number. If for
7044 some reason other than non-existence or emptiness, which it checks, it
7045 could not run './exim', it was installing it with an empty version number,
7046 i.e. as "exim-". This error state is now caught, and the installation is
7049 38. An argument was missing from the function that creates an error message
7050 when Exim fails to connect to the socket for saslauthd authentication.
7051 This could cause Exim to crash, or give a corrupted message.
7053 39. Added isip, isip4, and isip6 to ${if conditions.
7055 40. The ACL variables $acl_xx are now saved with the message, and can be
7056 accessed later in routers, transports, and filters.
7058 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7059 strings in the file are not string-expanded.
7061 42. If a MAIL command specified a SIZE value that was too large to fit into an
7062 int variable, the check against message_size_limit failed. Such values are
7063 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7064 day this will have to be increased, but I don't think I want to be around
7065 when emails are that large.
7072 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7073 iconv() is not standard in FreeBSD.
7075 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7076 IPv6 enabled. The observed symptom was a segmentation fault on return from
7077 the function os_common_find_running_interfaces() in src/os.c.
7079 3. In the check_special_case() function in daemon.c I had used "errno" as an
7080 argument name, which causes warnings on some systems. This was basically a
7081 typo, since it was named "eno" in the comments!
7083 4. The code that waits for the clock to tick (at a resolution of some fraction
7084 of a second) so as to ensure message-id uniqueness was always waiting for
7085 at least one whole tick, when it could have waited for less. [This is
7086 almost certainly not relevant at current processor speeds, where it is
7087 unlikely to ever wait at all. But we try to future-proof.]
7089 5. The function that sleeps for a time interval that includes fractions of a
7090 second contained a race. It did not block SIGALRM between setting the
7091 timer, and suspending (a couple of lines later). If the interval was short
7092 and the sigsuspend() was delayed until after it had expired, the suspension
7093 never ended. On busy systems this could lead to processes getting stuck for
7096 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7097 process, before it forks any delivery processes. The open lookup caching
7098 mechanism meant that the open file or database connection was passed into
7099 the delivery process. The problem was that delivery processes always tidy
7100 up cached lookup data. This could cause a problem for the next delivery
7101 process started by the queue runner, because the external queue runner
7102 process does not know about the closure. So the next delivery process
7103 still has data in the lookup cache. In the case of a file lookup, there was
7104 no problem because closing a file descriptor in a subprocess doesn't affect
7105 the parent. However, if the lookup was caching a connection to a database,
7106 the connection was closed, and the second delivery process was likely to
7107 see errors such as "PGSQL: query failed: server closed the connection
7108 unexpectedly". The problem has been fixed by closing all cached lookups
7109 in a queue runner before running a delivery process.
7111 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7112 seem to have the "const" qualifier which it has on other OS. I've
7115 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7116 *of the same type* (client or server) with the same public name that an
7117 error should be diagnosed.
7119 9. When Exim looked up a host name for an IP address, but failed to find the
7120 original IP address when looking up the host name (a safety check), it
7121 output the message "<ip address> does not match any IP for NULL", which was
7122 confusing, to say the least. The bug was that the host name should have
7123 appeared instead of "NULL".
7125 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7126 user that is built into the binary, and the -C or -D options is used, root
7127 privilege is dropped before the configuration file is read. In addition,
7128 logging is switched to stderr instead of the normal log files. If the
7129 configuration then re-defines the Exim user, the unprivileged environment
7130 is probably not what is expected, so Exim logs a panic warning message (but
7133 However, if deliver_drop_privilege is set, the unprivileged state may well
7134 be exactly what is intended, so the warning has been cut out in that case,
7135 and Exim is allowed to try to write to its normal log files.
7141 1. smtp_return_error_details was not giving details for temporary sender
7142 or receiver verification errors.
7144 2. Diagnose a configuration error if two authenticators have the same public
7147 3. Exim used not to create the message log file for a message until the first
7148 delivery attempt. This could be confusing when incoming messages were held
7149 for policy or load reasons. The message log file is now created at the time
7150 the message is received, and an initial "Received" line is written to it.
7152 4. The automatically generated man page for command line options had a minor
7153 bug that caused no ill effects; however, a more serious problem was that
7154 the procedure for building the man page automatically didn't always
7155 operate. Consequently, release 4.20 contains an out-of-date version. This
7156 shouldn't happen again.
7158 5. When building Exim with embedded Perl support, the script that builds the
7159 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7160 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7162 6. The freeze_tell option was not being used for messages that were frozen on
7163 arrival, either by an ACL or by local_scan().
7165 7. Added the smtp_incomplete_transaction log selector.
7167 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7168 was accepting AUTH without a new EHLO.
7170 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7171 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7174 10. Logging of TCP/IP connections (when configured) now happens in the main
7175 daemon process instead of the child process, so that the TCP/IP connection
7176 count is more accurate (but it can never be perfect).
7178 11. The use of "drop" in a nested ACL was not being handled correctly in the
7179 outer ACL. Now, if condition failure induced by the nested "drop" causes
7180 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7181 or "require"), the connection is dropped.
7183 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7184 that yield "discard" can now be used with an "accept" or a "discard" verb,
7185 but an error is generated for any others (because I can't see a useful way
7186 to define what should happen).
7188 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7189 are now processed in the same way as lines in the Exim configuration file.
7190 In particular, continuation lines are supported.
7192 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7194 15. Added -ti meaning -t -i.
7196 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7197 domains, and warn by logging if others are found.
7199 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7200 returned from GIFCONF when Exim is trying to find the list of interfaces on
7201 a host. The code in os.c has been modified to copy each ifreq to an aligned
7202 structure in all cases.
7204 Also, in some cases, the returned ifreq's were being copied to a 'struct
7205 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7206 means the last couple of bytes of an IPv6 address could be chopped if the
7207 ifreq contained only a normal sockaddr (14 bytes storage).
7209 18. Named domain lists were not supported in the hosts_treat_as_local option.
7210 An entry such as +xxxx was not recognized, and was treated as a literal
7213 19. Ensure that header lines added by a DATA ACL are included in the reject log
7214 if the ACL subsequently rejects the message.
7216 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7217 MD5 (which is deprecated).
7219 21. When testing a filter file using -bf, Exim was writing a message when it
7220 took the sender from a "From " line in the message, but it was not doing so
7221 when it took $return_path from a Return-Path: header line. It now does.
7223 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7224 with a valid header line field name (a series of printing characters
7225 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7227 23. Changed "disc" in the source to "disk" to conform to the documentation and
7228 the book and for uniformity.
7230 24. Ignore Sendmail's -Ooption=value command line item.
7232 25. When execve() failed while trying to run a command in a pipe transport,
7233 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7234 could be confused with a return value of 69 from the command itself. This
7235 has been changed to 127, the value the shell returns if it is asked to run
7236 a non-existent command. The wording for the related log line suggests a
7237 non-existent command as the problem.
7239 26. If received_header_text expands to an empty string, do not add a Received:
7240 header line to the message. (Well, it adds a token one on the spool, but
7241 marks it "old" so that it doesn't get used or transmitted.)
7243 27. Installed eximstats 1.28 (addition of -nt option).
7245 28. There was no check for failure on the call to getsockname() in the daemon
7246 code. This can fail if there is a shortage of resources on the system, with
7247 ENOMEM, for example. A temporary error is now given on failure.
7249 29. Contrary to the C standard, it seems that in some environments, the
7250 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7251 program. Exim now does this explicitly; it affects the formatting of
7252 timestamps using strftime().
7254 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7255 complaints. I've now initialized all the variables, to avoid this.
7257 32. Header lines added by a system filter were not being "seen" during
7258 transport-time rewrites.
7260 33. The info_callback() function passed to OpenSSL is set up with type void
7261 (*)(SSL *, int, int), as described somewhere. However, when calling the
7262 function (actually a macro) that sets it up, the type void(*)() is
7263 expected. I've put in a cast to prevent warnings from picky compilers.
7265 34. If a DNS black list lookup found a CNAME record, but there were no A
7266 records associated with the domain it pointed at, Exim crashed.
7268 35. If a DNS black list lookup returned more than one A record, Exim ignored
7269 all but the first. It now scans all returned addresses if a particular IP
7270 value is being sought. In this situation, the contents of the
7271 $dnslist_value variable are a list of all the addresses, separated by a
7274 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7275 to accept a host name and all its aliases if the forward lookup for any of
7276 them yielded the IP address of the incoming connection. Now it accepts only
7277 those names whose forward lookup yields the correct IP address. Any other
7278 names are discarded. This closes a loophole whereby a rogue DNS
7279 administrator could create reverse DNS records to break through a
7280 wildcarded host restriction in an ACL.
7282 37. If a user filter or a system filter that ran in a subprocess used any of
7283 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7284 the wrong values were passed to the pipe command ($thisaddress had the
7285 value of $0, $0 had the value of $1, etc). This bug was introduced by
7286 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7288 38. Improved the line breaking for long SMTP error messages from ACLs.
7289 Previously, if there was no break point between 40 and 75 characters, Exim
7290 left the rest of the message alone. Two changes have been made: (a) I've
7291 reduced the minimum length to 35 characters; (b) if it can't find a break
7292 point between 35 and 75 characters, it looks ahead and uses the first one
7293 that it finds. This may give the occasional overlong line, but at least the
7294 remaining text gets split now.
7296 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7297 file descriptors might be low, and that setting 1000 would always raise it.
7298 It turns out that in some environments, the limit is already over 1000 and
7299 that lowering it causes trouble. So now Exim takes care not to decrease it.
7301 40. When delivering a message, the value of $return_path is set to $sender_
7302 address at the start of routing (routers may change the value). By an
7303 oversight, this default was not being set up when an address was tested by
7304 -bt or -bv, which affected the outcome if any router or filter referred to
7307 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7308 log only when "message" or "log_message" are set. However, if one of the
7309 conditions was an address verification, or a call to a nested ACL, the
7310 messages generated by the underlying test were being passed through. This
7311 no longer happens. The underlying message is available in $acl_verify_
7312 message for both "message" and "log_message" expansions, so it can be
7313 passed through if needed.
7315 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7316 new expansion $bh_ to give the encoded byte string without charset
7317 translation. Translation happens only if iconv() is available; HAVE_ICONV
7318 indicates this at build time. HEADERS_CHARSET gives the charset to
7319 translate to; headers_charset can change it in the configuration, and
7320 "headers charset" can change it in an individual filter file.
7322 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7323 that creates RFC 2047 encoded "words" labels them as that charset instead
7324 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7325 expansion operator; (ii) when Exim creates a From: line for a local
7326 message; (iii) when a header line is rewritten to include a "phrase" part.
7328 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7329 buggy, causing it to skip the first lines of messages whose message ID
7330 ended in 'D'. This would not have bitten before Exim release 4.14, because
7331 message IDs were unlikely to end in 'D' before then. The effect was to have
7332 incorrect size information for certain domains.
7334 45. #include "config.h" was missing at the start of the crypt16.c module. This
7335 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7338 46. If there was a timeout during a "random" callout check, Exim treated it as
7339 a failure of the random address, and carried on sending RSET and the real
7340 address. If the delay was just some slowness somewhere, the response to the
7341 original RCPT would be taken as a response to RSET and so on, causing
7342 mayhem of various kinds.
7344 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7345 when I implemented it. It didn't allow for the fact that some option values
7346 may legitimately be negative (e.g. size_addition), and it didn't even do
7347 the right test for positive values.
7349 48. Domain names in DNS records are case-independent. Exim always looks them up
7350 in lower case. Some resolvers return domain names in exactly the case they
7351 appear in the zone file, that is, they may contain uppercase letters. Not
7352 all resolvers do this - some return always lower case. Exim was treating a
7353 change of case by a resolver as a change of domain, similar to a widening
7354 of a domain abbreviation. This triggered its re-routing code and so it was
7355 trying to route what was effectively the same domain again. This normally
7356 caused routing to fail (because the router wouldn't handle the domain
7357 twice). Now Exim checks for this case specially, and just changes the
7358 casing of the domain that it ultimately uses when it transmits the message
7361 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7364 50. If a filter generated a file delivery with a non-absolute name (possible if
7365 no home directory exists for the router), the forbid_file option was not
7368 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7369 the existing equality matching.
7371 52. Exim was using ints instead of ino_t variables in some places where it was
7372 dealing with inode numbers.
7374 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7375 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7376 called TMPDIR, and if it finds it is different, it changes its value.
7378 54. The smtp_printf() function is now made available to local_scan() so
7379 additional output lines can be written before returning. There is also an
7380 smtp_fflush() function to enable the detection of a dropped connection.
7381 The variables smtp_input and smtp_batched_input are exported to
7384 55. Changed the default runtime configuration: the message "Unknown user"
7385 has been removed from the ACL, and instead placed on the localuser router,
7386 using the cannot_route_message feature. This means that any verification
7387 failures that generate their own messages won't get overridden. Similarly,
7388 the "Unrouteable address" message that was in the ACL for unverifiable
7389 relay addresses has also been removed.
7391 56. Added hosts_avoid_esmtp to the smtp transport.
7393 57. The exicyclog script was not checking for the esoteric option
7394 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7395 will work only if exicyclog is run under the appropriate euid.
7397 58. Following a discussion on the list, the rules by which Exim recognises line
7398 endings on incoming messages have been changed. The -dropcr and drop_cr
7399 options are now no-ops, retained only for backwards compatibility. The
7400 following line terminators are recognized: LF CRLF CR. However, special
7401 processing applies to CR:
7403 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7404 nor a local message in the state where . is a terminator.
7406 (ii) If a bare CR is encountered in a header line, an extra space is added
7407 after the line terminator so as not to end the header. The reasoning
7408 behind this is that bare CRs in header lines are most likely either
7409 to be mistakes, or people trying to play silly games.
7411 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7412 was being incorrectly given as 18 bytes larger than it should have been.
7413 This is a VOB (very old bug).
7415 60. This may never have affected anything current, but just in case it has:
7416 When the local host is found other than at the start of a list of hosts,
7417 the local host, those with the same MX, and any that follow, are discarded.
7418 When the list in question was part of a longer list of hosts, the following
7419 hosts (not currently being processed) were also being discarded. This no
7420 longer happens. I'm not sure if this situation could ever has previously
7423 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7426 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7427 header. This is something that is recommended in a new Internet Draft, and
7428 is something that is documented as being done by Sendmail. There are two
7429 possible values. For messages generated by the autoreply transport, Exim
7432 Auto-Submitted: auto-replied
7434 whereas for all other generated messages (e.g. bounces) it adds
7436 Auto-Submitted: auto-generated
7438 63. The "personal" condition in filters now includes a test for the
7439 Auto-Submitted: header. If it contains the string "auto-" the message it
7440 not considered personal.
7442 64. Added rcpt_include_affixes as a generic transport option.
7444 65. Added queue_only_override (default true).
7446 66. Added the syslog_duplication option.
7448 67. If what should have been the first header line of a message consisted of
7449 a space followed by a colon, Exim was mis-interpreting it as a header line.
7450 It isn't of course - it is syntactically invalid and should therefore be
7451 treated as the start of the message body. The misbehaviour could have
7452 caused a number of strange effects, including loss of data in subsequent
7453 header lines, and spool format errors.
7455 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7456 client host had authenticated. This control can now be exercised by an ACL
7457 for more flexibility.
7459 69. By default, callouts do not happen when testing with -bh. There is now a
7460 variant, -bhc, which does actually run the callout code, including
7461 consulting and updating the callout cache.
7463 70. Added support for saslauthd authentication, courtesy of Alexander
7466 71. If statvfs() failed on the spool or log directories while checking their
7467 size for availability, Exim confusingly gave the error "space shortage".
7468 Furthermore, in debugging mode it crashed with a floating point exception.
7469 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7470 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7471 serious problem, Exim now writes to the main and panic logs when this
7472 happens, with details of the failure. It then refuses to accept the
7473 incoming message, giving the message "spool directory problem" or "log
7474 directory problem" with a 421 code for SMTP messages.
7476 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7477 0, 1, and 2 exist, because some OS complain for execs without them (see
7478 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7479 descriptors. However, the code omitted to check that the open succeeded,
7480 causing mysterious errors if for some reason the permissions on /dev/null
7481 got screwed. Now Exim writes a message to the main and panic logs, and
7482 bombs out if it can't open /dev/null.
7484 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7485 interact so that it is all more flexible. It is supposed to remain
7486 backwards compatible. Also added extra_local_interfaces.
7488 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7489 to bomb out with an assertion failure - to the client this appears as a
7490 connection drop. This problem occurs in the part of the code that was taken
7491 from the Samba project. Fortunately, the assertion is in a very simple
7492 function, so I have fixed this by reproducing the function inline in the
7493 one place where it is called, and arranging for authentication to fail
7494 instead of killing the process with assert().
7496 75. The SPA client code was not working when the server requested OEM rather
7497 than Unicode encoding.
7499 76. Added code to make require_files with a specific uid setting more usable in
7500 the case where statting the file as root fails - usually a non-root-mounted
7501 NFS file system. When this happens and the failure is EACCES, Exim now
7502 forks a subprocess and does the per-uid checking as the relevant uid.
7504 77. Added process_log_path.
7506 78. If log_file_path was not explicitly set, a setting of check_log_space or
7507 check_log_inodes was ignored.
7509 79. If a space check for the spool or log partitions fails, the incident is now
7510 logged. Of course, in the latter case the data may get lost...
7512 80. Added the %p formatting code to string_format() so that it can be used to
7513 print addresses in debug_print(). Adjusted all the address printing in the
7514 debugging in store.c to use %p rather than %d.
7516 81. There was a concern that a line of code in smtp_in.c could overflow a
7517 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7518 initially expressed, the concern was not well-founded, because trailing
7519 spaces are removed early. However, if the trailing spaces were followed by
7520 a NULL, they did not get removed, so the overflow was possible. Two fixes
7523 (a) I re-wrote the offending code in a cleaner fashion.
7524 (b) If an incoming SMTP command contains a NULL character, it is rejected
7527 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7528 runs initgroups(), so that if the Exim user is in any additional groups,
7529 they will be used during message reception.
7535 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob