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.
145 JH/01 Remove code calling the customisable local_scan function, unless a new
146 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
148 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
149 non-signal-safe functions being used.
151 JH/03 Bug 2269: When presented with a received message having a stupidly large
152 number of DKIM-Signature headers, disable DKIM verification to avoid
153 a resource-consumption attack. The limit is set at twenty.
155 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
156 report of oldest_pass in ${authres } in consequence, and separate out
157 some descriptions of reasons for verification fail.
159 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
160 files in the spool were present and unlocked. A queue-runner could spot
161 them, resulting in a duplicate delivery. Fix that by doing the unlock
162 after the unlink. Investigation by Tim Stewart. Take the opportunity to
163 add more error-checking on spoolfile handling while that code is being
166 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
167 No known attacks, no CVE, this is defensive hardening.
169 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
170 a queue-runner could start a delivery while other operations were ongoing.
171 Cutthrough delivery was a common victim, resulting in duplicate delivery.
172 Found and investigated by Tim Stewart. Fix by using the open message data
173 file handle rather than opening another, and not locally closing it (which
174 releases a lock) for that case, while creating the temporary .eml format
175 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
177 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
178 $sender_verify_failure/$recipient_verify_failure to "random".
180 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
183 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
184 Previously this would segfault.
186 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
189 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
190 like zero, since the resolver should be doing this for us, But we need one
191 as a CNAME but no MX presence gets the CNAME returned; we need to check
192 that doesn't point to an MX to declare it "no MX returned" rather than
193 "error, loop". A new main option is added so the older capability of
194 following some limited number of chain links is maintained.
196 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
198 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
199 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
200 now-deprecated earlier definitions used only specified the range up to TLS
201 1.2 (in the older-version library docs).
203 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
205 JH/15 Rework TLS client-side context management. Stop using a global, and
206 explicitly pass a context around. This enables future use of TLS for
207 connections to service-daemons (eg. malware scanning) while a client smtp
208 connection is using TLS; with cutthrough connections this is quite likely.
210 JH/16 Fix ARC verification to do AS checks in reverse order.
212 JH/17 Support a "tls" option on the ${readsocket } expansion item.
214 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
215 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
216 Previously the "utf8" would be re-prepended for every additional message.
218 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
219 Previously thery were accepted, resulting in issues when attempting to
220 forward messages to a non-supporting MTA.
222 PP/02 Let -n work with printing macros too, not just options.
224 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
225 one parent address was copied, and bogus data was used at delivery-logging
226 time. Either a crash (after delivery) or bogus log data could result.
227 Discovery and analysis by Tim Stewart.
229 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
230 Previously if the string ended mid-character, we did not insert the
231 promised '?' replacement.
233 PP/04 Documentation: current string operators work on bytes, not codepoints.
235 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
236 should pack well giving a smaller memory footprint so better caching and
237 therefore performance. Group the declarations where this can't be done so
238 that the byte-sized flag variables are not interspersed among pointer
239 variables, giving a better chance of good packing by the compiler.
241 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
242 non-null, to avoid issues with sites running BATV. Previously reports were
243 sent with an empty envelope sender so looked like bounces.
245 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
246 The ignore_error flag wasn't being returned from the filter subprocess so
247 was not set for later routers. Investigation and fix by Matthias Kurz.
249 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
250 and a msg:complete for the whole, when a message is manually removed using
251 -Mrm. Developement by Matthias Kurz, hacked on by JH.
253 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
254 a "Gnu special" function, asprintf() in the DB utility binary builds; I
255 hope that is portable enough.
257 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
258 requiring a known-CA anchor certificate; make it now rely entirely on the
259 TLSA as an anchor. Checking the name on the leaf cert against the name
260 on the A-record for the host is still done for TA (but not for EE mode).
262 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
263 included in delivery lines for non-proxied connections, when compiled with
264 SUPPORT_SOCKS and running with proxy logging enabled.
266 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
267 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
268 move the existing event to fire before the normal logging of message
269 failure so that custom logging is bracketed by normal logging.
271 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
272 msg:fail:internal event. Developement by Matthias Kurz.
274 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
275 far too small for todays use of crypto signatures stored there. Go all
276 the way to the max DNS message size of 64kB, even though this might be
277 overmuch for IOT constrained device use.
279 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
280 copy a string over itself. The library routine is documented as not
281 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
283 JH/32 For main options check_spool_space and check_inode_space, where the
284 platform supports 64b integers, support more than the previous 2^31 kB
285 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
286 the previous G, M, k.
288 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
289 $authenticated_fail_id variable on authentication failure. Previously
292 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
293 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
294 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
295 GNUTLS_SEC_PARAM_MEDIUM.
297 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
298 side. Previously we would continue as if no SNI had been received.
300 JH/36 Harden the handling of string-lists. When a list consisted of a sole
301 "<" character, which should be a list-separator specification, we walked
302 off past the nul-terimation.
304 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
305 causes) even when the retry time is not yet met. Previously they were
306 not, meaning that when (say) an account was over-quota and temp-rejecting,
307 and multiple senders' messages were queued, only one sender would get
308 notified on each configured delay_warning cycle.
310 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
312 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
313 cipher-suites, an error can be left on the stack even for a succeeding
314 accept; this results in impossible error messages when a later operation
317 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
318 return error codes indicating retry. Under TLS1.3 this becomes required.
320 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
321 it only wrote the new authenticators, resulting in a lack of tracking of
322 peer changes of ESMTP extensions until the next cache flush.
324 JH/41 Fix the loop reading a message header line to check for integer overflow,
325 and more-often against header_maxsize. Previously a crafted message could
326 induce a crash of the recive process; now the message is cleanly rejected.
328 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
329 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
335 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
336 When redis_servers is set to a list of > 1 element, and the Redis servers
337 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
338 case of MOVED into a DEFER case instead, thus moving the query onto the
339 next server in the list. For a cluster of N elements, all N servers must
340 be defined in redis_servers.
342 GF/02 Catch and remove uninitialized value warning in exiqsumm
343 Check for existence of @ARGV before looking at $ARGV[0]
345 JH/01 Replace the store_release() internal interface with store_newblock(),
346 which internalises the check required to safely use the old one, plus
347 the allocate and data copy operations duplicated in both (!) of the
348 extant use locations.
350 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
351 modifier. This matches the restriction on the commandline.
353 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
354 Previously only the last row was returned.
356 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
357 we assumed that tags in the header were well-formed, and parsed the
358 element content after inspecting only the first char of the tag.
359 Assumptions at that stage could crash the receive process on malformed
362 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
363 While running the DKIM ACL we operate on the Permanent memory pool so that
364 variables created with "set" persist to the DATA ACL. Also (at any time)
365 DNS lookups that fail create cache records using the Permanent pool. But
366 expansions release any allocations made on the current pool - so a dnsdb
367 lookup expansion done in the DKIM ACL releases the memory used for the
368 DNS negative-cache, and bad things result. Solution is to switch to the
369 Main pool for expansions.
370 While we're in that code, add checks on the DNS cache during store_reset,
371 active in the testsuite.
372 Problem spotted, and debugging aided, by Wolfgang Breyha.
374 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
375 When none of the hosts presented to a transport match an already-open
376 connection, close it and proceed with the list. Previously we would
377 queue the message. Spotted by Lena with Yahoo, probably involving
380 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
381 Previously a spurious "250 OK id=" response was appended to the proper
384 JH/08 The "support for" informational output now, which built with Content
385 Scanning support, has a line for the malware scanner interfaces compiled
386 in. Interface can be individually included or not at build time.
388 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
389 by the template makefile "src/EDITME". The "STREAM" support for an older
390 ClamAV interface method is removed.
392 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
393 rows affected is given instead).
395 JH/11 The runtime Berkeley DB library version is now additionally output by
396 "exim -d -bV". Previously only the compile-time version was shown.
398 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
399 SMTP connection. Previously, when one had more receipients than the
400 first, an abortive onward connection was made. Move to full support for
401 multiple onward connections in sequence, handling cutthrough connection
402 for all multi-message initiating connections.
404 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
405 routers. Previously, a multi-recipient message would fail to match the
406 onward-connection opened for the first recipient, and cause its closure.
408 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
409 a timeout on read on a GnuTLS initiating connection, resulting in the
410 initiating connection being dropped. This mattered most when the callout
411 was marked defer_ok. Fix to keep the two timeout-detection methods
414 JH/15 Relax results from ACL control request to enable cutthrough, in
415 unsupported situations, from error to silently (except under debug)
416 ignoring. This covers use with PRDR, frozen messages, queue-only and
419 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
421 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
422 metadata, resulting in a crash in free().
424 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
425 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
426 Broken also in d185889f4, with init system revamp.
428 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
429 Previously we abruptly closed the connection after reading a malware-
430 found indication; now we go on to read the "scan ok" response line,
433 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
434 ACL. Previously, a crash would result.
436 JH/19 Speed up macro lookups during configuration file read, by skipping non-
437 macro text after a replacement (previously it was only once per line) and
438 by skipping builtin macros when searching for an uppercase lead character.
440 JH/20 DANE support moved from Experimental to mainline. The Makefile control
441 for the build is renamed.
443 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
444 was allocated for every new TLS startup, meaning one per message. Fix
445 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
447 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
448 reported the original. Fix to report (as far as possible) the ACL
449 result replacing the original.
451 JH/23 Fix memory leak during multi-message connections using STARTTLS under
452 OpenSSL. Certificate information is loaded for every new TLS startup,
453 and the resources needed to be freed.
455 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
457 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
460 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
461 DATA response info to the (existing) per-recipient response info for
462 the "C=" log element. It can have useful tracking info from the
463 destination system. Patch from Simon Arlott.
465 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
466 length value. Previously this would segfault.
468 HS/02 Support Avast multiline protoocol, this allows passing flags to
469 newer versions of the scanner.
471 JH/28 Ensure that variables possibly set during message acceptance are marked
472 dead before release of memory in the daemon loop. This stops complaints
473 about them when the debug_store option is enabled. Discovered specifically
474 for sender_rate_period, but applies to a whole set of variables.
475 Do the same for the queue-runner and queue-list loops, for variables set
476 from spool message files. Do the same for the SMTP per-message loop, for
477 certain variables indirectly set in ACL operations.
479 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
480 as a multi-recipient message from a mailinglist manager). The coding had
481 an arbitrary cutoff number of characters while checking for more input;
482 enforced by writing a NUL into the buffer. This corrupted long / fast
483 input. The problem was exposed more widely when more pipelineing of SMTP
484 responses was introduced, and one Exim system was feeding another.
485 The symptom is log complaints of SMTP syntax error (NUL chars) on the
486 receiving system, and refused recipients seen by the sending system
487 (propating to people being dropped from mailing lists).
488 Discovered and pinpointed by David Carter.
490 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
491 replaced by the ${authresults } expansion.
493 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
495 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
496 allows proper process termination in container environments.
498 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
499 Previously the "final dot" had a newline after it; ensure it is CR,LF.
501 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
502 and "err_perm", deprecated since 4.83 when the RFC-defined words
503 "temperror" and "permerror" were introduced.
505 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
506 transport-filters or DKIM-signing. The restriction was lost in the
507 consolidation of verify-callout and delivery SMTP handling.
508 Extend the restriction to also cover ARC-signing.
510 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
511 in defer=pass mode supply a 450 to the initiator. Previously the message
514 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
515 tls_require_ciphers is used as before.
517 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
518 "pass_unscanned". Only tmpfails from the scanner are written to
519 the paniclog, as they may require admin intervention (permission
520 denied, license issues). Other scanner errors (like decompression
521 bombs) do not cause a paniclog entry.
523 JH/36 Fix reinitialisation of DKIM logging variable between messages.
524 Previously it was possible to log spurious information in receive log
527 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
528 triggered odd behaviour from Outlook Express clients.
530 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
533 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
534 since the IETF WG has not yet settled on that versus the original
535 "bare" representation.
537 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
538 Previously the millisecond value corrupted the output.
539 Fix also for syslog_pid=no and log_selector +pid, for which the pid
540 corrupted the output.
546 JH/01 Rework error string handling in TLS interface so that the caller in
547 more cases is responsible for logging. This permits library-sourced
548 string to be attached to addresses during delivery, and collapses
549 pairs of long lines into single ones.
551 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
552 during configuration. Wildcards are allowed and expanded.
554 JH/02 Rework error string handling in DKIM to pass more info back to callers.
555 This permits better logging.
557 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
558 do not close it down and have the child transport start it up again on
559 the passed-on TCP connection. Instead, proxy the child (and any
560 subsequent ones) for TLS via a unix-domain socket channel. Logging is
561 affected: the continued delivery log lines do not have any DNSSEC, TLS
562 Certificate or OCSP information. TLS cipher information is still logged.
564 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
565 identical IP addresses on different listening ports. Will also affect
568 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
569 add noisy ifdef guards to special-case this sillyness.
570 Patch from Bernd Kuhls.
572 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
573 than 255 are no longer allowed.
575 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
576 Disable the session-cache too, which might reduce our load. Since we
577 currrectly use a new context for every connection, both as server and
578 client, there is no benefit for these.
579 GnuTLS appears to not support tickets server-side by default (we don't
580 call gnutls_session_ticket_enable_server()) but client side is enabled
581 by default on recent versions (3.1.3 +) unless the PFS priority string
584 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
585 <https://reproducible-builds.org/specs/source-date-epoch/>.
587 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
588 the check for any unsuccessful recipients did not notice the limit, and
589 erroneously found still-pending ones.
591 JH/08 Pipeline CHUNKING command and data together, on kernels that support
592 MSG_MORE. Only in-clear (not on TLS connections).
594 JH/09 Avoid using a temporary file during transport using dkim. Unless a
595 transport-filter is involved we can buffer the headers in memory for
596 creating the signature, and read the spool data file once for the
597 signature and again for transmission.
599 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
600 4.77 as the kernel support then wasn't solid, having issues in 64bit
601 mode. Now, it's been long enough. Add support for FreeBSD also.
603 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
604 case where the routing stage had gathered several addresses to send to
605 a host before calling the transport for the first, we previously failed
606 to close down TLS in the old transport process before passing the TCP
607 connection to the new process. The new one sent a STARTTLS command
608 which naturally failed, giving a failed delivery and bloating the retry
609 database. Investigation and fix prototype from Wolfgang Breyha.
611 JH/12 Fix check on SMTP command input synchronisation. Previously there were
612 false-negatives in the check that the sender had not preempted a response
613 or prompt from Exim (running as a server), due to that code's lack of
614 awareness of the SMTP input buffering.
616 PP/04 Add commandline_checks_require_admin option.
617 Exim drops privileges sanely, various checks such as -be aren't a
618 security problem, as long as you trust local users with access to their
619 own account. When invoked by services which pass untrusted data to
620 Exim, this might be an issue. Set this option in main configuration
621 AND make fixes to the calling application, such as using `--` to stop
624 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
625 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
626 responses to those, into a single TLS record each way (this usually means
627 a single packet). As a side issue, smtp_enforce_sync now works on TLS
630 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
631 affects you only if you're dancing at the edge of the param size limits.
632 If you are, and this message makes sense to you, then: raise the
633 configured limit or use OpenSSL 1.1. Nothing we can do for older
636 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
637 cmdline element to get the documented default one. Previously it was
640 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
641 are unsuable for non-message-specific reasons - eg. connection timeout,
642 banner-time rejection.
644 JH/16 Fix logging of delivery remote port, when specified by router, under
647 PP/06 Repair manualroute's ability to take options in any order, even if one
648 is the name of a transport.
651 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
653 JH/17 Change the list-building routines interface to use the expanding-string
654 triplet model, for better allocation and copying behaviour.
656 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
657 Previously it was constructed the first time a possibly-matching string
658 was met in the configuration file input during startup; now it is done
661 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
662 compatible one, to avoid the (poorly documented) possibility of a config
663 file in the working directory redirecting the DB files, possibly correpting
664 some existing file. CVE-2017-10140 assigned for BDB.
666 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
667 cache-hot. Previously, although the result was properly cached, the
668 initial verify call returned a defer.
670 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
671 the main verify for receipient in uncached-mode.
673 JH/22 Retire historical build files to an "unsupported" subdir. These are
674 defined as "ones for which we have no current evidence of testing".
676 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
677 if present. Previously it was ignored.
679 JH/24 Start using specified-initialisers in C structure init coding. This is
680 a C99 feature (it's 2017, so now considered safe).
682 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
683 if was a fixed-sized field and bitmask ops via macros; it is now more
686 PP/07 GitHub PR 56: Apply MariaDB build fix.
687 Patch provided by Jaroslav Škarvada.
689 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
690 during Coverity cleanups [4.87 JH/47]
691 Diagnosis and fix provided by Michael Fischer v. Mollard.
693 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
694 the right size to place the terminating semicolon on its own folded
695 line, the header hash was calculated to an incorrect value thanks to
696 the (relaxed) space the fold became.
698 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
699 and confused the parent.
701 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
702 which could crash as a result. This could lead to undeliverable messages.
704 JH/28 Logging: "next input sent too soon" now shows where input was truncated
707 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
708 matters on fast-turnover and PID-randomising systems, which were getting
709 out-of-order delivery.
711 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
712 a possibly-overlapping copy. The symptom was that "Remote host closed
713 connection in response to HELO" was logged instead of the actual 4xx
716 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
717 Previously only that bufferd was discarded, resulting in SYMTP command
720 JH/32 DKIM: when a message has multiple signatures matching an identity given
721 in dkim_verify_signers, run the dkim acl once for each. Previously only
722 one run was done. Bug 2189.
724 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
725 "panic the current process" to "deliberately defer". The panic log is
726 still written with the problem list name; the mail and reject logs now
727 get a temp-reject line for the message that was being handled, saying
728 something like "domains check lookup or other defer". The SMTP 451
729 message is still "Temporary local problem".
731 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
732 A crafted sequence of BDAT commands could result in in-use memory beeing
733 freed. CVE-2017-16943.
735 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
736 from SMTP input. Previously it was always done; now only done for DATA
737 and not BDAT commands. CVE-2017-16944.
739 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
740 to the message (such as an overlong header line). Previously this was
741 not done and we did not exit BDAT mode. Followon from the previous item
742 though a different problem.
748 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
749 than -2003 did; needs libidn2 in addition to libidn.
751 JH/02 The path option on a pipe transport is now expanded before use.
753 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
754 Patch provided by "Björn", documentation fix added too.
756 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
757 missing a wire-to-host endian conversion.
759 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
760 close after a BDAT command line could be taken as a following command,
761 giving a synch failure. Fix by only checking for synch immediately
762 before acknowledging the chunk.
764 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
765 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
766 macro. Patches provided by Josh Soref.
768 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
769 Previously we did not; the RFC seems ambiguous and VRFY is not listed
770 by IANA as a service extension. However, John Klensin suggests that we
773 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
774 the dkim code may be unix-mode line endings rather than smtp wire-format
775 CRLF, so prepend a CR to any bare LF.
777 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
778 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
780 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
781 the first were themselves being wrongly included in the feed into dkim
782 processing; with most chunk sizes in use this resulted in an incorrect
783 body hash calculated value.
785 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
786 DKIM signature block, for verification. Although advised against by
787 standards it is specifically not ruled illegal.
789 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
791 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
792 missing a body hash (the bh= tag).
794 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
795 It seems that HAProxy sends the Proxy Protocol information in clear and
796 only then does a TLS startup, so do the same.
798 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
799 TCP connections (such as for Spamd) unless the daemon successfully set
800 Fast Open mode on its listening sockets. This fixes breakage seen on
801 too-old kernels or those not configured for Fast Open, at the cost of
802 requiring both directions being enabled for TFO, and TFO never being used
803 by non-daemon-related Exim processes.
805 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
806 endings, at least on the first header line. Try to canonify any that get
807 past that check, despite the cost.
809 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
810 now limited to an arbitrary five deep, while parsing addresses with the
811 strip_excess_angle_brackets option enabled.
813 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
814 instead leave the unprompted TLS handshake in socket buffer for the
815 TLS library to consume.
817 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
819 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
821 JH/16 Drop variables when they go out of scope. Memory management drops a whole
822 region in one operation, for speed, and this leaves assigned pointers
823 dangling. Add checks run only under the testsuite which checks all
824 variables at a store-reset and panics on a dangling pointer; add code
825 explicitly nulling out all the variables discovered. Fixes one known
826 bug: a transport crash, where a dangling pointer for $sending_ip_address
827 originally assigned in a verify callout, is re-used.
829 PP/06 Drop '.' from @INC in various Perl scripts.
831 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
833 PP/08 Reduce a number of compilation warnings under clang; building with
834 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
835 should be warning-free.
837 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
839 HS/01 Fix portability problems introduced by PP/08 for platforms where
840 realloc(NULL) is not equivalent to malloc() [SunOS et al].
842 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
843 chunk. This allows us to accept broken chunked messages. We need a more
844 general solution here.
846 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
847 already-broken messages in the queue.
849 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
851 JH/19 Fix reference counting bug in routing-generated-address tracking.
857 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
858 supports it and a size is available (ie. the sending peer gave us one).
860 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
861 years of being deprecated). The replacements are the ACLs
862 acl_smtp_mime and acl_not_smtp_mime.
864 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
865 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
866 or even an in-clear connection were permitted. Now, if the host lookup
867 was dnssec and dane was requested then the host is only used if the
868 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
869 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
870 if one fails this test.
871 This means that a poorly-configured remote DNS will make it incommunicado;
872 but it protects against a DNS-interception attack on it.
874 JH/04 Bug 1810: make continued-use of an open smtp transport connection
875 non-noisy when a race steals the message being considered.
877 JH/05 If main configuration option tls_certificate is unset, generate a
878 self-signed certificate for inbound TLS connections.
880 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
881 in rewrites and routers.
883 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
884 and logged a warning sing 4.83; now they are a configuration file error.
886 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
887 (lacking @domain). Apply the same qualification processing as RCPT.
889 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
891 JH/10 Support ${sha256:} applied to a string (as well as the previous
894 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
895 a cutthrough deliver is pending, as we always want to make a connection.
896 This also avoids re-routing the message when later placing the cutthrough
897 connection after a verify cache hit.
898 Do not update it with the verify result either.
900 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
901 when routing results in more than one destination address.
903 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
904 signing (which inhibits the cutthrough capability). Previously only
905 the presence of an option was tested; now an expansion evaluating as
906 empty is permissible (obviously it should depend only on data available
907 when the cutthrough connection is made).
909 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
910 the relevant preceding SMTP command did not note the pipelining mode.
912 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
913 Previously they were not counted.
915 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
916 as one having no matching records. Previously we deferred the message
917 that needed the lookup.
919 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
920 distinguished as "(=".
922 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
923 for missing MX records. Previously it only worked for missing A records.
925 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
927 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
928 after the data-go-ahead and data-ack. Patch from Jason Betts.
930 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
931 even for a "none" policy. Patch from Tony Meyer.
933 JH/22 Fix continued use of a connection for further deliveries. If a port was
934 specified by a router, it must also match for the delivery to be
937 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
938 When one of the recipients of a message was unsuitable for the connection
939 (has no matching addresses), we lost track of needing to mark it
940 deferred. As a result mail would be lost.
942 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
944 JH/25 Decoding ACL controls is now done using a binary search; the source code
945 takes up less space and should be simpler to maintain. Merge the ACL
946 condition decode tables also, with similar effect.
948 JH/26 Fix problem with one_time used on a redirect router which returned the
949 parent address unchanged. A retry would see the parent address marked as
950 delivered, so not attempt the (identical) child. As a result mail would
953 JH/27 Fix a possible security hole, wherein a process operating with the Exim
954 UID can gain a root shell. Credit to http://www.halfdog.net/ for
955 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
958 JH/28 Enable {spool,log} filesystem space and inode checks as default.
959 Main config options check_{log,spool}_{inodes,space} are now
960 100 inodes, 10MB unless set otherwise in the configuration.
962 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
963 Previously it only applied to the main-section connection policy
966 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
968 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
969 by me. Added RFC7919 DH primes as an alternative.
971 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
972 are not in the system include path.
974 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
975 GnuTLS, when a session startup failed (eg because the client disconnected)
976 Exim did stdio operations after fclose. This was exposed by a recent
977 change which nulled out the file handle after the fclose.
979 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
980 signed directly by the cert-signing cert, rather than an intermediate
981 OCSP-signing cert. This is the model used by LetsEncrypt.
983 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
985 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
986 an incoming connection.
988 HS/02 Bug 1802: Do not half-close the connection after sending a request
991 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
992 fallback to "prime256v1".
994 JH/34 SECURITY: Use proper copy of DATA command in error message.
995 Could leak key material. Remotely exploitable. CVE-2016-9963.
1001 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1002 and 3.4.4 - once the server is enabled to respond to an OCSP request
1003 it does even when not requested, resulting in a stapling non-aware
1004 client dropping the TLS connection.
1006 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1007 support variable-length bit vectors. No functional change.
1009 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1010 The I= interface field on outgoing lines is now after the H= remote
1011 host field, same as incoming lines. There is a separate
1012 outgoing_interface log selector which allows you to disable the
1015 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1016 If not running log_selector +smtp_connection the mainlog would be held
1017 open indefinitely after a "too many connections" event, including to a
1018 deleted file after a log rotate. Leave the per net connection logging
1019 leaving it open for efficiency as that will be quickly detected by the
1020 check on the next write.
1022 HS/01 Bug 1671: Fix post transport crash.
1023 Processing the wait-<transport> messages could crash the delivery
1024 process if the message IDs didn't exist for some reason. When
1025 using 'split_spool_directory=yes' the construction of the spool
1026 file name failed already, exposing the same netto behaviour.
1028 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1029 mime_regex ACL conditions.
1031 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1032 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1033 to HELO, local diagnostic string.
1035 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1036 log line to debug. Even when configured with a tls authenticator many
1037 client connections are expected to not authenticate in this way, so
1038 an authenticate fail is not an error.
1040 HS/02 Add the Exim version string to the process info. This way exiwhat
1041 gives some more detail about the running daemon.
1043 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1044 matter for fast-change records such as DNSBLs.
1046 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1047 retry record, even if constant. There may be multiple transports with
1048 different interface settings and the retry behaviour needs to be kept
1051 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1053 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1055 JH/10 Bug 840: fix log_defer_output option of pipe transport
1057 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1058 hosts_randomize. This matters a lot when combined with mua_wrapper.
1060 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1061 ${quote_pgsql:<string>} operator.
1063 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1065 JH/14 Change header file name for freeradius-client. Relevant if compiling
1066 with Radius support; from the Gentoo tree and checked under Fedora.
1068 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1070 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1071 option transport_filter, meaning no filtering.
1073 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1075 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1076 defaults to "*" (all hosts). The variable is now available when not built
1077 with TLS, default unset, mainly to enable keeping the testsuite sane.
1078 If a server certificate is not supplied (via tls_certificate) an error is
1079 logged, and clients will find TLS connections fail on startup. Presumably
1080 they will retry in-clear.
1081 Packagers of Exim are strongly encouraged to create a server certificate
1082 at installation time.
1084 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1085 with the $config_file variable.
1087 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1088 in transport context, after the attempt, and per-recipient. The latter type
1089 is per host attempted. The event data is the error message, and the errno
1090 information encodes the lookup type (A vs. MX) used for the (first) host,
1091 and the trailing two digits of the smtp 4xx response.
1093 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1094 to write to mainlog (or rejectlog, paniclog) in the window between file
1095 creation and permissions/ownership being changed. Particularly affects
1096 installations where exicyclog is run as root, rather than exim user;
1097 result is that the running daemon panics and dies.
1099 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1101 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1102 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1103 "pri" and "weight". Note that the previous implicit priority given by the
1104 list order is no longer honoured.
1106 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1107 for DKIM processing.
1109 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1110 by defining SUPPORT_SOCKS.
1112 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1113 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1114 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1115 variables are renamed to proxy_{local,external}_{address,port}.
1117 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1118 for a build by defining SUPPORT_I18N
1120 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1121 of the query string, and make ${quote_redis:} do that quoting.
1123 JH/29 Move Events support from Experimental to mainline, enabled by default
1124 and removable for a build by defining DISABLE_EVENT.
1126 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1128 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1129 cached by the daemon.
1131 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1132 by defining LOOKUP_REDIS. The libhiredis library is required.
1134 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1135 keys are given for lookup.
1137 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1138 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1139 only supported when built with TLS support. The PolarSSL SHA routines
1140 are still used when the TLS library is too old for convenient support.
1142 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1143 openssl_options), for security. OpenSSL forces this from version 1.1.0
1144 server-side so match that on older versions.
1146 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1147 allocation for $value could be released as the expansion processing
1148 concluded, but leaving the global pointer active for it.
1150 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1151 and to use the domains and local_parts ACL conditions.
1153 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1154 incorrectly not doubled on cutthrough transmission, hence seen as a
1155 body-termination at the receiving system - resulting in truncated mails.
1156 Commonly the sender saw a TCP-level error, and retransmitted the message
1157 via the normal store-and-forward channel. This could result in duplicates
1158 received - but deduplicating mailstores were liable to retain only the
1159 initial truncated version.
1161 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1163 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1165 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1166 we're in there, support oversigning also; bug 1309.
1168 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1170 HS/04 Add support for keep_environment and add_environment options.
1172 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1173 either intentional arithmetic overflow during PRNG, or testing config-
1176 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1177 delivery resulted in actual delivery. Cancel cutthrough before DATA
1180 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1181 rejects a recipient: pass the reject to the originator.
1183 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1184 Many were false-positives and ignorable, but it's worth fixing the
1187 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1188 for the new environment-manipulation done at startup. Move the routines
1189 from being local to tls.c to being global via the os.c file.
1191 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1192 an extract embedded as result-arg for a map, the first arg for extract
1193 is unavailable so we cannot tell if this is a numbered or keyed
1194 extraction. Accept either.
1200 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1203 JH/02 The smtp transport option "multi_domain" is now expanded.
1205 JH/03 The smtp transport now requests PRDR by default, if the server offers
1208 JH/04 Certificate name checking on server certificates, when exim is a client,
1209 is now done by default. The transport option tls_verify_cert_hostnames
1210 can be used to disable this per-host. The build option
1211 EXPERIMENTAL_CERTNAMES is withdrawn.
1213 JH/05 The value of the tls_verify_certificates smtp transport and main options
1214 default to the word "system" to access the system default CA bundle.
1215 For GnuTLS, only version 3.0.20 or later.
1217 JH/06 Verification of the server certificate for a TLS connection is now tried
1218 (but not required) by default. The verification status is now logged by
1219 default, for both outbound TLS and client-certificate supplying inbound
1222 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1225 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1226 Status Notification (bounce) messages are now MIME format per RFC 3464.
1227 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1228 under the control of the dsn_advertise_hosts option, and routers may
1229 have a dsn_lasthop option.
1231 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1232 default, modifiable by a malware= option. The list separator for
1233 the options can now be changed in the usual way. Bug 68.
1235 JH/10 The smtp_receive_timeout main option is now expanded before use.
1237 JH/11 The incoming_interface log option now also enables logging of the
1238 local interface on delivery outgoing connections.
1240 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1241 if the interface and destination host and port all match.
1243 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1246 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1247 Patch from Andrew Lewis.
1249 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1250 now supports optional time-restrictions, weighting, and priority
1251 modifiers per server. Patch originally by <rommer@active.by>.
1253 JH/16 The spamd_address main option now supports a mixed list of local
1254 and remote servers. Remote servers can be IPv6 addresses, and
1255 specify a port-range.
1257 JH/17 Bug 68: The spamd_address main option now supports an optional
1258 timeout value per server.
1260 JH/18 Bug 1581: Router and transport options headers_add/remove can
1261 now have the list separator specified.
1263 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1266 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1269 JH/21 Support for the A6 type of dns record is withdrawn.
1271 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1272 rather than the verbs used.
1274 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1275 from 255 to 1024 chars.
1277 JH/24 Verification callouts now attempt to use TLS by default.
1279 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1280 are generic router options now. The defaults didn't change.
1282 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1283 Original patch from Alexander Shikoff, worked over by JH.
1285 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1286 files if ZCAT_COMMAND is not executable.
1288 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1290 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1292 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1293 Normally benign, it bites when the pair was led to by a CNAME;
1294 modern usage is to not canonicalize the domain to a CNAME target
1295 (and we were inconsistent anyway for A-only vs AAAA+A).
1297 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1299 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1300 when evaluating $sender_host_dnssec.
1302 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1303 $sender_helo_dnssec variable.
1305 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1307 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1309 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1311 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1312 documented as working, but never had. Support all but $spam_report.
1314 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1315 added for tls authenticator.
1317 HS/03 Add perl_taintmode main config option
1323 TL/01 When running the test suite, the README says that variables such as
1324 no_msglog_check are global and can be placed anywhere in a specific
1325 test's script, however it was observed that placement needed to be near
1326 the beginning for it to behave that way. Changed the runtest perl
1327 script to read through the entire script once to detect and set these
1328 variables, reset to the beginning of the script, and then run through
1329 the script parsing/test process like normal.
1331 TL/02 The BSD's have an arc4random API. One of the functions to induce
1332 adding randomness was arc4random_stir(), but it has been removed in
1333 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1334 function when detected.
1336 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1337 cause callback expansion.
1339 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1340 syntax errors in an expansion can be treated as a string instead of
1341 logging or causing an error, due to the internal use of bool_lax
1342 instead of bool when processing it.
1344 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1345 server certificates when making smtp deliveries.
1347 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1349 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1351 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1353 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1354 Merged patch from Sebastian Wiedenroth.
1356 JH/05 Fix results-pipe from transport process. Several recipients, combined
1357 with certificate use, exposed issues where response data items split
1358 over buffer boundaries were not parsed properly. This eventually
1359 resulted in duplicates being sent. This issue only became common enough
1360 to notice due to the introduction of connection certificate information,
1361 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1363 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1364 size buffer was used, resulting in syntax errors when an expansion
1367 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1368 version 3.3.6 or later.
1370 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1371 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1372 both become "event_action", the variables become $event_name, $event_data
1373 and $event_defer_errno. There is a new variable $verify_mode, usable in
1374 routers, transports and related events. The tls:cert event is now also
1375 raised for inbound connections, if the main configuration event_action
1378 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1379 early OCSP support, but no stapling (appears to be less than 1.0.0).
1381 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1382 server certificate names available under the smtp transport option
1383 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1386 JH/10 Time-related extraction expansions from certificates now use the main
1387 option "timezone" setting for output formatting, and are consistent
1388 between OpenSSL and GnuTLS compilations. Bug 1541.
1390 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1391 encoded parameter in the incoming message. Bug 1558.
1393 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1394 include certificate info, eximon was claiming there were spoolfile
1397 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1399 JH/14 Log delivery-related information more consistently, using the sequence
1400 "H=<name> [<ip>]" wherever possible.
1402 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1403 are problematic for Debian distribution, omit them from the release
1406 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1408 JH/16 Fix string representation of time values on 64bit time_t architectures.
1411 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1412 output list separator was used.
1417 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1418 checkers that were complaining about end of non-void function with no
1421 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1422 This was a regression introduced in 4.83 by another bugfix.
1424 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1426 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1427 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1433 TF/01 Correctly close the server side of TLS when forking for delivery.
1435 When a message was received over SMTP with TLS, Exim failed to clear up
1436 the incoming connection properly after forking off the child process to
1437 deliver the message. In some situations the subsequent outgoing
1438 delivery connection happened to have the same fd number as the incoming
1439 connection previously had. Exim would try to use TLS and fail, logging
1440 a "Bad file descriptor" error.
1442 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1443 utilities have not been installed.
1445 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1446 temporary space as the ACL may create new global variables.
1448 TL/01 LDAP support uses per connection or global context settings, depending
1449 upon the detected version of the libraries at build time.
1451 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1452 to extract and use the src ip:port in logging and expansions as if it
1453 were a direct connection from the outside internet. PPv2 support was
1454 updated based on HAProxy spec change in May 2014.
1456 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1458 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1459 Properly escape header and check for NULL return.
1461 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1464 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1466 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1467 characters in header names, implemented as a verify condition.
1468 Contributed by Michael Fischer v. Mollard.
1470 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1471 results permerror and temperror. Previous values are deprecated but
1472 still accepted. In a future release, err_perm and err_temp will be
1473 completely removed, which will be a backward incompatibility if the
1474 ACL tests for either of these two old results. Patch contributed by
1475 user bes-internal on the mailing list.
1477 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1479 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1480 selectors, in both main and reject logs.
1482 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1485 JH/07 Add malware type "sock" for talking to simple daemon.
1487 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1489 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1490 routers/transports under cutthrough routing.
1492 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1493 numbers. Touch up "bool" conditional to keep the same definition.
1495 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1497 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1499 JH/12 Expand items in router/transport headers_add or headers_remove lists
1500 individually rather than the list as a whole. Bug 1452.
1502 Required for reasonable handling of multiple headers_ options when
1503 they may be empty; requires that headers_remove items with embedded
1504 colons must have them doubled (or the list-separator changed).
1506 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1507 view the policy declared in the DMARC record. Currently, $dmarc_status
1508 is a combined value of both the record presence and the result of the
1511 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1513 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1514 dnslookup router and the smtp transport (applying to the forward
1517 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1518 of ldap servers used for a specific lookup. Patch provided by Heiko
1521 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1522 New variable $lookup_dnssec_authenticated for observability.
1524 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1525 Patch submitted by Lars Timman.
1527 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1529 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1530 Requires trusted mode and valid format message id, aborts otherwise.
1531 Patch contributed by Heiko Schlichting.
1533 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1534 certextract with support for various fields. Bug 1358.
1536 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1537 is requested by default, modifiable by smtp transport option
1540 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1541 operate on certificate variables to give certificate fingerprints
1542 Also new ${sha256:cert_variable}.
1544 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1546 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1547 Christian Aistleitner.
1549 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1551 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1552 file. Patch from Wolfgang Breyha.
1554 JH/25 Expand the coverage of the delivery $host and $host_address to
1555 client authenticators run in verify callout. Bug 1476.
1557 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1558 align with daemon_smtp_ports. Bug 72.
1560 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1561 support and error reporting did not work properly.
1563 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1564 and is readable. Patch from Andrew Colin Kissa.
1566 TL/14 Enhance documentation of ${run expansion and how it parses the
1567 commandline after expansion, particularly in the case when an
1568 unquoted variable expansion results in an empty value.
1570 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1572 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1573 by renaming away from it. Addresses GH issue 15, reported by
1576 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1578 TL/15 SECURITY: prevent double expansion in math comparison functions
1579 (can expand unsanitized data). Not remotely exploitable.
1586 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1588 PP/02 Make -n do something, by making it not do something.
1589 When combined with -bP, the name of an option is not output.
1591 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1594 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1595 $sender_host_name and config options to manage this, and basic check
1598 PP/05 DSCP support for outbound connections and control modifier for inbound.
1600 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1601 (Only plugin which currently uses this is kerberos4, which nobody should
1602 be using, but we should make it available and other future plugins might
1603 conceivably use it, even though it would break NAT; stuff *should* be
1604 using channel bindings instead).
1606 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1607 name; added for Sendmail compatibility; requires admin caller.
1608 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1609 just ignore it); requires trusted caller.
1610 Also parse but ignore: -Ac -Am -X<logfile>
1613 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1615 TL/02 Add +smtp_confirmation as a default logging option.
1617 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1618 Patch by Magnus Holmgren from 2007-02-20.
1620 TL/04 Bugzilla 1281 - Spec typo.
1621 Bugzilla 1283 - Spec typo.
1622 Bugzilla 1290 - Spec grammar fixes.
1624 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1626 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1628 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1629 mailing list by by Dmitry Isaikin.
1631 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1633 JH/02 Support "G" suffix to numbers in ${if comparisons.
1635 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1637 NM/01 Bugzilla 1197 - Spec typo
1638 Bugzilla 1196 - Spec examples corrections
1640 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1642 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1643 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1646 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1647 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1649 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1650 "acl {{name}{arg}...}", and optional args on acl condition
1653 JH/05 Permit multiple router/transport headers_add/remove lines.
1655 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1657 JH/07 Avoid using a waiting database for a single-message-only transport.
1658 Performance patch from Paul Fisher. Bugzilla 1262.
1660 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1663 JH/09 Add $headers_added variable, with content from use of ACL modifier
1664 add_header (but not yet added to the message). Bugzilla 199.
1666 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1667 Pulled from Bugzilla 817 by Wolfgang Breyha.
1669 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1671 (nb: this is the same fix as in Exim 4.80.1)
1673 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1676 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1679 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1681 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1682 advertise SMTP AUTH mechanism to us, instead of a generic
1683 protocol violation error. Also, make Exim more robust to bad
1684 data from the Dovecot auth socket.
1686 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1688 When a queue runner is handling a message, Exim first routes the
1689 recipient addresses, during which it prunes them based on the retry
1690 hints database. After that it attempts to deliver the message to
1691 any remaining recipients. It then updates the hints database using
1694 So if a recipient address works intermittently, it can get repeatedly
1695 deferred at routing time. The retry hints record remains fresh so the
1696 address never reaches the final cutoff time.
1698 This is a fairly common occurrence when a user is bumping up against
1699 their storage quota. Exim had some logic in its local delivery code
1700 to deal with this. However it did not apply to per-recipient defers
1701 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1703 This change adds a proper retry rule check during routing so that the
1704 final cutoff time is checked against the message's age. We only do
1705 this check if there is an address retry record and there is not a
1706 domain retry record; this implies that previous attempts to handle
1707 the address had the retry_use_local_parts option turned on. We use
1708 this as an approximation for the destination being like a local
1709 delivery, as in LMTP.
1711 I suspect this new check makes the old local delivery cutoff check
1712 redundant, but I have not verified this so I left the code in place.
1714 TF/02 Correct gecos expansion when From: is a prefix of the username.
1716 Test 0254 submits a message to Exim with the header
1720 When I ran the test suite under the user fanf2, Exim expanded
1721 the header to contain my full name, whereas it should have added
1722 a Resent-Sender: header. It erroneously treats any prefix of the
1723 username as equal to the username.
1725 This change corrects that bug.
1727 GF/01 DCC debug and logging tidyup
1728 Error conditions log to paniclog rather than rejectlog.
1729 Debug lines prefixed by "DCC: " to remove any ambiguity.
1731 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1733 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1734 Bug spotted by Jeremy Harris; was flawed since initial commit.
1735 Would have resulted in OCSP responses post-SNI triggering an Exim
1736 NULL dereference and crash.
1738 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1740 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1741 Bug detection, analysis and fix by Samuel Thibault.
1742 Bugzilla 1331, Debian bug #698092.
1744 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1746 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1747 Server implementation by Todd Lyons, client by JH.
1748 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1749 config variable "prdr_enable" controls whether the server
1750 advertises the facility. If the client requests PRDR a new
1751 acl_data_smtp_prdr ACL is called once for each recipient, after
1752 the body content is received and before the acl_smtp_data ACL.
1753 The client is controlled by both of: a hosts_try_prdr option
1754 on the smtp transport, and the server advertisement.
1755 Default client logging of deliveries and rejections involving
1756 PRDR are flagged with the string "PRDR".
1758 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1759 fclose(). Diagnosis by Todd Lyons.
1761 PP/17 Update configure.default to handle IPv6 localhost better.
1762 Patch by Alain Williams (plus minor tweaks).
1765 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1766 This is now consistent with GnuTLS, and is now documented: the
1767 previous undocumented portable approach to treating the option as
1768 unset was to force an expansion failure. That still works, and
1769 an empty string is now equivalent.
1771 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1772 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1773 not performing validation itself.
1775 PP/20 Added force_command boolean option to pipe transport.
1776 Patch from Nick Koston, of cPanel Inc.
1778 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1781 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1783 PP/21 Fix eximon continuous updating with timestamped log-files.
1784 Broken in a format-string cleanup in 4.80, missed when I repaired the
1785 other false fix of the same issue.
1786 Report and fix from Heiko Schlichting.
1789 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1790 Report from Prashanth Katuri.
1792 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1793 It's SecureTransport, so affects any MacOS clients which use the
1794 system-integrated TLS libraries, including email clients.
1796 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1797 using a MIME ACL for non-SMTP local injection.
1798 Report and assistance in diagnosis by Warren Baker.
1800 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1802 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1804 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1805 last id that failed so it may be referenced in subsequent ACL's.
1807 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1810 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1811 ldap library initialization, allowing self-signed CA's to be
1812 used. Also properly sets require_cert option later in code by
1813 using NULL (global ldap config) instead of ldap handle (per
1814 session). Bug diagnosis and testing by alxgomz.
1816 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1817 the src/util/ subdirectory.
1819 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1820 renamed to Transport Post Delivery Action by Jeremy Harris, as
1823 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1824 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1825 redis_servers = needs to be configured which will be used by the redis
1826 lookup. Patch from Warren Baker, of The Packet Hub.
1828 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1830 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1831 hostname or reverse DNS when processing a host list. Used suggestions
1832 from multiple comments on this bug.
1834 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1836 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1837 Missed a few lines, added it to make the runtest require no keyboard
1840 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1841 contains upper case chars. Make router use caseful_local_part.
1843 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1844 support when GnuTLS has been built with p11-kit.
1850 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1852 This, or similar/improved, will also be change PP/11 of 4.82.
1858 PP/01 Handle short writes when writing local log-files.
1859 In practice, only affects FreeBSD (8 onwards).
1860 Bugzilla 1053, with thanks to Dmitry Isaikin.
1862 NM/01 Bugzilla 949 - Documentation tweak
1864 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1867 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1869 PP/02 Implemented gsasl authenticator.
1871 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1873 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1874 `pkg-config foo` for cflags/libs.
1876 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1877 with rest of GSASL and with heimdal_gssapi.
1879 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1880 `pkg-config foo` for cflags/libs for the TLS implementation.
1882 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1883 properties get this fed in as external SSF. A number of robustness
1884 and debugging improvements to the cyrus_sasl authenticator.
1886 PP/08 cyrus_sasl server now expands the server_realm option.
1888 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1889 Patch by Jeremy Harris.
1891 PP/10 Added dbmjz lookup type.
1893 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1895 PP/12 MAIL args handles TAB as well as SP, for better interop with
1896 non-compliant senders.
1897 Analysis and variant patch by Todd Lyons.
1899 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1900 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1901 Patch from Dirk Mueller <dmueller@suse.com>
1903 PP/13 tls_peerdn now print-escaped for spool files.
1904 Observed some $tls_peerdn in wild which contained \n, which resulted
1905 in spool file corruption.
1907 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1908 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1909 or write after TLS renegotiation, which otherwise led to messages
1912 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1913 as a tracking header (ie: a signed header comes before the signature).
1914 Patch from Wolfgang Breyha.
1916 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1917 comma-sep list; embedded commas doubled.
1919 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1921 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1923 Report and patch from Dmitry Banschikov.
1925 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1926 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1927 failures. We appear to now support TLS1.1+ with Exim.
1929 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1930 lets Exim select keys and certificates based upon TLS SNI from client.
1931 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1932 before an outbound SMTP session. New log_selector, +tls_sni.
1934 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1935 NULL dereference. Report and patch from Alun Jones.
1937 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1938 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1939 Not seeing resolver debug output on NetBSD, but suspect this is a
1940 resolver implementation change.
1942 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1943 Left warnings. Added "eximon gdb" invocation mode.
1945 PP/21 Defaulting "accept_8bitmime" to true, not false.
1947 PP/22 Added -bw for inetd wait mode support.
1949 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1950 locate the relevant includes and libraries. Made this the default.
1952 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1953 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1955 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1956 This may cause build issues on older platforms.
1958 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1959 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1960 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1961 Added SNI support via GnuTLS too.
1962 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1964 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1966 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1967 Applied second patch from Janne, implementing suggestion to default
1968 multiple-strings-in-record handling to match SPF spec.
1970 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1972 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1973 read-only, out of scope).
1974 Patch from Wolfgang Breyha, report from Stuart Northfield.
1976 PP/29 Fix three issues highlighted by clang analyser static analysis.
1977 Only crash-plausible issue would require the Cambridge-specific
1978 iplookup router and a misconfiguration.
1979 Report from Marcin Mirosław.
1981 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1983 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1984 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1985 As part of this, removing so much warning spew let me fix some minor
1986 real issues in debug logging.
1988 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1989 assignment on my part. Fixed.
1991 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1992 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1993 Janne Snabb (who went above and beyond: thank you).
1995 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1996 string otherwise requires a connection and a bunch more work and it's
1997 relatively easy to get wrong. Should also expose TLS library linkage
2000 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2001 64-bit ${eval} (JH/03).
2003 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2004 GNU libc to support some of the 64-bit stuff, should not lead to
2005 conflicts. Defined before os.h is pulled in, so if a given platform
2006 needs to override this, it can.
2008 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2009 protection layer was required, which is not implemented.
2010 Bugzilla 1254, patch from Wolfgang Breyha.
2012 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2013 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2014 tls_dhparam take prime identifiers. Also unbreak combination of
2015 OpenSSL+DH_params+TLSSNI.
2017 PP/39 Disable SSLv2 by default in OpenSSL support.
2023 PP/01 Solaris build fix for Oracle's LDAP libraries.
2024 Bugzilla 1109, patch from Stephen Usher.
2026 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2028 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2031 TF/02 Fix a couple more cases where we did not log the error message
2032 when unlink() failed. See also change 4.74-TF/03.
2034 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2035 lock up or crash if it happened to be inside a call to libc when it
2036 got a SIGUSR1 from exiwhat.
2038 The SIGUSR1 handler appends the current process status to the process
2039 log which is later printed by exiwhat. It used to use the general
2040 purpose logging code to do this, but several functions it calls are
2041 not safe for signals.
2043 The new output code in the SIGUSR1 handler is specific to the process
2044 log, and simple enough that it's easy to inspect for signal safety.
2045 Removing some special cases also simplifies the general logging code.
2046 Removing the spurious timestamps from the process log simplifies
2049 TF/04 Improved ratelimit ACL condition.
2051 The /noupdate option has been deprecated in favour of /readonly which
2052 has clearer semantics. The /leaky, /strict, and /readonly update modes
2053 are mutually exclusive. The update mode is no longer included in the
2054 database key; it just determines when the database is updated. (This
2055 means that when you upgrade Exim will forget old rate measurements.)
2057 Exim now checks that the per_* options are used with an update mode that
2058 makes sense for the current ACL. For example, when Exim is processing a
2059 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2060 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2061 must specify per_mail/readonly. If you omit the update mode it defaults to
2062 /leaky where that makes sense (as before) or /readonly where required.
2064 The /noupdate option is now undocumented but still supported for
2065 backwards compatibility. It is equivalent to /readonly except that in
2066 ACLs where /readonly is required you may specify /leaky/noupdate or
2067 /strict/noupdate which are treated the same as /readonly.
2069 A useful new feature is the /count= option. This is a generalization
2070 of the per_byte option, so that you can measure the throughput of other
2071 aggregate values. For example, the per_byte option is now equivalent
2072 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2074 The per_rcpt option has been generalized using the /count= mechanism
2075 (though it's more complicated than the per_byte equivalence). When it is
2076 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2077 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2078 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2079 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2080 Note that using per_rcpt with a non-readonly update mode in more than
2081 one ACL will cause the recipients to be double-counted. (The per_mail
2082 and per_byte options don't have this problem.)
2084 The handling of very low rates has changed slightly. If the computed rate
2085 is less than the event's count (usually one) then this event is the first
2086 after a long gap. In this case the rate is set to the same as this event's
2087 count, so that the first message of a spam run is counted properly.
2089 The major new feature is a mechanism for counting the rate of unique
2090 events. The new per_addr option counts the number of different
2091 recipients that someone has sent messages to in the last time period. It
2092 behaves like per_rcpt if all the recipient addresses are different, but
2093 duplicate recipient addresses do not increase the measured rate. Like
2094 the /count= option this is a general mechanism, so the per_addr option
2095 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2096 example, measure the rate that a client uses different sender addresses
2097 with the options per_mail/unique=$sender_address. There are further
2098 details in the main documentation.
2100 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2102 TF/06 Removed a few PCRE remnants.
2104 TF/07 Automatically extract Exim's version number from tags in the git
2105 repository when doing development or release builds.
2107 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2108 Bugzilla 879. Patch from Paul Fisher.
2110 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2111 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2114 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2116 PP/05 Variable $av_failed, true if the AV scanner deferred.
2117 Bugzilla 1078. Patch from John Horne.
2119 PP/06 Stop make process more reliably on build failure.
2120 Bugzilla 1087. Patch from Heiko Schlittermann.
2122 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2123 Bugzilla 1089. Patch from Heiko Schlittermann.
2125 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2126 Bugzilla 1131. Patch from Holger Weiß.
2128 PP/09 Handle IPv6 addresses with SPF.
2129 Bugzilla 860. Patch from Wolfgang Breyha.
2131 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2133 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2136 PP/11 match_* no longer expand right-hand-side by default.
2137 New compile-time build option, EXPAND_LISTMATCH_RHS.
2138 New expansion conditions, "inlist", "inlisti".
2140 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2142 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2144 PP/14 fix log_write() format string regression from TF/03.
2145 Bugzilla 1152. Patch from Dmitry Isaikin.
2151 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2153 PP/02 Harmonised TLS library version reporting; only show if debugging.
2154 Layout now matches that introduced for other libraries in 4.74 PP/03.
2156 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2158 PP/04 New "dns_use_edns0" global option.
2160 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2163 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2164 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2166 TK/01 Updated PolarSSL code to 0.14.2.
2167 Bugzilla 1097. Patch from Andreas Metzler.
2169 PP/07 Catch divide-by-zero in ${eval:...}.
2170 Fixes bugzilla 1102.
2172 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2175 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2176 format-string attack -- SECURITY: remote arbitrary code execution.
2178 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2179 time unintentionally subject to list matching rules, letting the header
2180 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2181 arbitrary string expansion). This allowed for information disclosure.
2183 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2184 INT_MIN/-1 -- value coerced to INT_MAX.
2190 NM/01 Workaround for PCRE version dependency in version reporting
2193 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2194 This fixes portability to compilers other than gcc, notably
2195 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2197 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2198 makefiles for portability to HP-UX and POSIX correctness.
2200 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2201 Also via indented variable definition in the Makefile.
2202 (Debugging by Oliver Heesakkers).
2204 PP/02 Restore caching of spamd results with expanded spamd_address.
2205 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2207 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2208 Improves build reliability. Fix from: Frank Elsner
2210 NM/02 Fix wide character breakage in the rfc2047 coding
2211 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2213 NM/03 Allow underscore in dnslist lookups
2214 Fixes bug 1026. Patch from Graeme Fowler
2216 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2217 Code patches from Adam Ciarcinski of NetBSD.
2219 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2222 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2223 is logged, to avoid truncation. Patch from John Horne.
2225 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2226 Patch from Jakob Hirsch.
2228 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2229 SQL string expansion failure details.
2230 Patch from Andrey Oktyabrski.
2232 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2233 Patch from Simon Arlott.
2235 PP/09 New lookups functionality failed to compile on old gcc which rejects
2236 extern declarations in function scope.
2237 Patch from Oliver Fleischmann
2239 PP/10 Use sig_atomic_t for flags set from signal handlers.
2240 Check getgroups() return and improve debugging.
2241 Fixed developed for diagnosis in bug 927 (which turned out to be
2244 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2245 Patch from Mark Zealey.
2247 PP/12 Bugzilla 1056: Improved spamd server selection.
2248 Patch from Mark Zealey.
2250 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2251 Based on patch from Heiko Schlittermann.
2253 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2254 Patch from Uwe Doering, sign-off by Michael Haardt.
2256 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2257 variable declaration deep within a block. Bug and patch from
2260 PP/15 lookups-Makefile IRIX compatibility coercion.
2262 PP/16 Make DISABLE_DKIM build knob functional.
2264 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2265 Patch by Simon Arlott
2267 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2268 variable argument macros. Our copy now differs from upstream.
2274 TF/01 Failure to get a lock on a hints database can have serious
2275 consequences so log it to the panic log.
2277 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2278 controlled using the smtp_confirmation log selector.
2280 TF/03 Include the error message when we fail to unlink a spool file.
2282 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2283 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2284 for maintaining out-of-tree patches for some time.
2286 PP/01 Bugzilla 139: Documentation and portability issues.
2287 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2288 Handle per-OS dynamic-module compilation flags.
2290 PP/02 Let /dev/null have normal permissions.
2291 The 4.73 fixes were a little too stringent and complained about the
2292 permissions on /dev/null. Exempt it from some checks.
2293 Reported by Andreas M. Kirchwitz.
2295 PP/03 Report version information for many libraries, including
2296 Exim version information for dynamically loaded libraries. Created
2297 version.h, now support a version extension string for distributors
2298 who patch heavily. Dynamic module ABI change.
2300 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2301 privilege escalation vulnerability whereby the Exim run-time user
2302 can cause root to append content of the attacker's choosing to
2305 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2308 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2309 If dropping privileges for untrusted macros, we disabled normal logging
2310 on the basis that it would fail; for the Exim run-time user, this is not
2311 the case, and it resulted in successful deliveries going unlogged.
2312 Fixed. Reported by Andreas Metzler.
2318 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2319 only prepend for the Resent-* case. Fixes regression introduced in
2320 Exim 4.70 by NM/22 for Bugzilla 607.
2322 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2323 increasing numbers of administrators be bitten by this.
2325 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2327 PP/03 Bugzilla 994: added openssl_options main configuration option.
2329 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2331 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2333 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2335 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2336 without a peer certificate, leading to a segfault because of an
2337 assumption that peers always have certificates. Be a little more
2338 paranoid. Problem reported by Martin Tscholak.
2340 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2341 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2342 NB: ClamAV planning to remove STREAM in "middle of 2010".
2343 CL also introduces -bmalware, various -d+acl logging additions and
2344 more caution in buffer sizes.
2346 PP/09 Implemented reverse_ip expansion operator.
2348 PP/10 Bugzilla 937: provide a "debug" ACL control.
2350 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2352 PP/12 Bugzilla 973: Implement --version.
2354 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2356 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2358 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2360 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2361 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2362 ignore trailing whitespace.
2364 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2366 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2369 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2370 Notification from Dr Andrew Aitchison.
2372 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2373 ExtendedDetectionInfo response format.
2374 Notification from John Horne.
2376 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2379 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2380 XSL and documented dependency on system catalogs, with examples of how
2383 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2386 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2387 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2388 configuration file which is writeable by the Exim user or group.
2390 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2391 of configuration files to cover files specified with the -C option if
2392 they are going to be used with root privileges, not just the default
2395 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2396 option (effectively making it always true).
2398 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2399 files to be used while preserving root privileges.
2401 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2402 that rogue child processes cannot use them.
2404 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2405 run-time user, instead of root.
2407 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2408 Exim run-time user without dropping privileges.
2410 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2411 result string, instead of calling string_vformat() twice with the same
2414 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2415 for other users. Others should always drop root privileges if they use
2416 -C on the command line, even for a whitelisted configure file.
2418 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2420 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2426 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2427 $data_path, and $header_path variables; fixed documentation bugs and
2430 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2431 exipick to access non-standard spools, including the "frozen" queue
2434 NM/01 Bugzilla 965: Support mysql stored procedures.
2435 Patch from Alain Williams
2437 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2439 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2440 Patch from Andreas Metzler
2442 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2443 Patch from Kirill Miazine
2445 NM/05 Bugzilla 671: Added umask to procmail example.
2447 JJ/03 installed exipick 20100323.0, fixing doc bug
2449 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2450 directory. Notification and patch from Dan Rosenberg.
2452 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2454 TK/02 Improve log output when DKIM signing operation fails.
2456 MH/01 Treat the transport option dkim_domain as a colon separated
2457 list, not as a single string, and sign the message with each element,
2458 omitting multiple occurences of the same signer.
2460 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2461 Bugzilla 985, 986. Patch by Simon Arlott
2463 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2464 Patch by Simon Arlott
2466 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2467 MBX locking. Notification from Dan Rosenberg.
2473 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2475 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2477 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2479 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2481 NM/04 Bugzilla 915: Flag broken perl installation during build.
2487 TK/01 Added patch by Johannes Berg that expands the main option
2488 "spamd_address" if it starts with a dollar sign.
2490 TK/02 Write list of recipients to X-Envelope-Sender header when building
2491 the mbox-format spool file for content scanning (suggested by Jakob
2494 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2495 (http://www.dcc-servers.net/) support via dccifd. Activated by
2496 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2498 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2499 by Mark Daniel Reidel <mr@df.eu>.
2501 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2502 When building exim an external PCRE library is now needed -
2503 PCRE is a system library on the majority of modern systems.
2504 See entry on PCRE_LIBS in EDITME file.
2506 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2507 conversation. Added nologin parameter to request.
2508 Patch contributed by Kirill Miazine.
2510 TF/01 Do not log submission mode rewrites if they do not change the address.
2512 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2514 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2515 log files in place. Contributed by Roberto Lima.
2517 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2519 TF/03 Bugzilla 615: When checking the local_parts router precondition
2520 after a local_part_suffix or local_part_prefix option, Exim now
2521 does not use the address's named list lookup cache, since this
2522 contains cached lookups for the whole local part.
2524 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2525 Robert Millan. Documentation is in experimental-spec.txt.
2527 TF/04 Bugzilla 668: Fix parallel build (make -j).
2529 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2531 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2532 Patch provided by Jan Srzednicki.
2534 TF/05 Leading white space used to be stripped from $spam_report which
2535 wrecked the formatting. Now it is preserved.
2537 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2538 that they are available at delivery time.
2540 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2542 TF/08 TLS error reporting now respects the incoming_interface and
2543 incoming_port log selectors.
2545 TF/09 Produce a more useful error message if an SMTP transport's hosts
2546 setting expands to an empty string.
2548 NM/06 Bugzilla 744: EXPN did not work under TLS.
2549 Patch provided by Phil Pennock.
2551 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2552 Patch provided by Richard Godbee.
2554 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2555 acl_smtp_notquit, added index entry.
2557 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2558 Patch provided by Eugene Bujak.
2560 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2561 accept(). Patch provided by Maxim Dounin.
2563 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2564 Patch provided by Phil Pennock.
2566 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2568 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2569 Patch provided by Brad "anomie" Jorsch.
2571 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2572 Patch provided by Dean Brooks.
2574 TK/05 Add native DKIM support (does not depend on external libraries).
2576 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2577 Patch provided by Graeme Fowler.
2579 NM/16 Bugzilla 851: Documentation example syntax fix.
2581 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2583 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2586 NM/19 Bugzilla 745: TLS version reporting.
2587 Patch provided by Phil Pennock.
2589 NM/20 Bugzilla 167: bool: condition support.
2590 Patch provided by Phil Pennock.
2592 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2593 clients. Patch provided by Phil Pennock.
2595 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2596 Patch provided by Brad "anomie" Jorsch.
2598 NM/23 Bugzilla 687: Fix misparses in eximstats.
2599 Patch provided by Heiko Schlittermann.
2601 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2602 Patch provided by Heiko Schlittermann.
2604 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2605 plus update to original patch.
2607 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2609 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2610 Patch provided by David Brownlee.
2612 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2614 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2616 NM/30 Bugzilla 888: TLS documentation bugfixes.
2618 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2620 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2621 Unlike the original bugzilla I have changed all shell scripts in src tree.
2623 NM/33 Bugzilla 898: Transport filter timeout fix.
2624 Patch by Todd Rinaldo.
2626 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2627 Patch by Serge Demonchaux.
2629 NM/35 Bugzilla 39: Base64 decode bug fixes.
2630 Patch by Jakob Hirsch.
2632 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2634 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2636 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2638 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2644 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2645 ALT-N's libdkim that I have put here:
2646 http://duncanthrax.net/exim-experimental/
2648 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2649 were called 'true' and it seems that C99 defines that as a reserved
2650 keyword to be used with 'bool' variable types. That means you could
2651 not include C99-style headers which use bools without triggering
2652 build errors in sieve.c.
2654 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2655 as mailq or other aliases. Changed the --help handling significantly
2656 to do whats expected. exim_usage() emits usage/help information.
2658 SC/01 Added the -bylocaldomain option to eximstats.
2660 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2662 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2664 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2670 PH/01 Another patch from the Sieve maintainer.
2672 PH/02 When an IPv6 address is converted to a string for single-key lookup
2673 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2674 dots are used instead of colons so that keys in lsearch files need not
2675 contain colons. This was done some time before quoting was made available
2676 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2677 (notated using the quote facility) so as to distinguish them from IPv4
2678 keys. This meant that lookups for IP addresses in host lists did not work
2679 for iplsearch lookups.
2681 This has been fixed by arranging for IPv6 addresses to be expressed with
2682 colons if the lookup type is iplsearch. This is not incompatible, because
2683 previously such lookups could never work.
2685 The situation is now rather anomalous, since one *can* have colons in
2686 ordinary lsearch keys. However, making the change in all cases is
2687 incompatible and would probably break a number of configurations.
2689 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2692 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2693 conversion specification without a maximum field width, thereby enabling
2694 a rogue spamd server to cause a buffer overflow. While nobody in their
2695 right mind would setup Exim to query an untrusted spamd server, an
2696 attacker that gains access to a server running spamd could potentially
2697 exploit this vulnerability to run arbitrary code as the Exim user.
2699 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2700 $primary_hostname instead of what libspf2 thinks the hosts name is.
2702 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2703 a directory entry by the name of the lookup key. Previously, if a
2704 symlink pointed to a non-existing file or a file in a directory that
2705 Exim lacked permissions to read, a lookup for a key matching that
2706 symlink would fail. Now it is enough that a matching directory entry
2707 exists, symlink or not. (Bugzilla 503.)
2709 PH/03 The body_linecount and body_zerocount variables are now exported in the
2712 PH/04 Added the $dnslist_matched variable.
2714 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2715 This means they are set thereafter only if the connection becomes
2718 PH/06 Added the client_condition to authenticators so that some can be skipped
2719 by clients under certain conditions.
2721 PH/07 The error message for a badly-placed control=no_multiline_responses left
2722 "_responses" off the end of the name.
2724 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2726 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2727 (without spaces) instead of just copying the configuration text.
2729 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2731 PH/11 Added $max_received_linelength.
2733 PH/12 Added +ignore_defer and +include_defer to host lists.
2735 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2736 way in which PCRE > 7.0 is built.
2738 PH/14 Implemented queue_only_load_latch.
2740 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2741 MAIL command. The effect was to mangle the value on 64-bit systems.
2743 PH/16 Another patch from the Sieve maintainer.
2745 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2747 PH/18 If a system quota error occurred while trying to create the file for
2748 a maildir delivery, the message "Mailbox is full" was not appended to the
2749 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2750 only to a quota excession during the actual writing of the file.
2752 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2753 characters?) which causes problems in log lines. The DN values are now
2754 passed through string_printing() before being added to log lines.
2756 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2757 and InterBase are left for another time.)
2759 PH/21 Added message_body_newlines option.
2761 PH/22 Guard against possible overflow in moan_check_errorcopy().
2763 PH/23 POSIX allows open() to be a macro; guard against that.
2765 PH/24 If the recipient of an error message contained an @ in the local part
2766 (suitably quoted, of course), incorrect values were put in $domain and
2767 $local_part during the evaluation of errors_copy.
2773 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2774 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2777 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2778 issue a MAIL command.
2780 PH/02 In an ACL statement such as
2782 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2784 if a client was not listed at all, or was listed with a value other than
2785 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2786 the condition was not true (as it should be), so access was not denied.
2787 The bug was that the ! inversion was incorrectly passed on to the second
2788 item. This has been fixed.
2790 PH/03 Added additional dnslists conditions == and =& which are different from
2791 = and & when the dns lookup returns more than one IP address.
2793 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2794 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2796 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2797 FSYNC, which compiles an option called disable_fsync that allows for
2798 bypassing fsync(). The documentation is heavily laced with warnings.
2800 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2802 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2803 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2804 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2805 including adding "make clean"; (3) Added -fPIC when compiling the test
2806 dynamically loaded module, to get rid of a warning.
2808 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2809 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2810 = 0s. The bug is otherwise harmless.
2812 PH/07 There was a bug in the dovecot authenticator such that the value of
2813 $auth1 could be overwritten, and so not correctly preserved, after a
2814 successful authentication. This usually meant that the value preserved by
2815 the server_setid option was incorrect.
2817 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2819 PH/09 Installed PCRE release 7.0.
2821 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2822 run for batched SMTP input. It is now run at the start of every message
2823 in the batch. While fixing this I discovered that the process information
2824 (output by running exiwhat) was not always getting set for -bs and -bS
2825 input. This is fixed, and it now also says "batched" for BSMTP.
2827 PH/11 Added control=no_pipelining.
2829 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2830 patch, slightly modified), and move the expansion of helo_data till after
2831 the connection is made in the smtp transport (so it can use these
2834 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2836 PH/14 Added log_selector = +pid.
2838 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2840 PH/16 Add ${if forany and ${if forall.
2842 PH/17 Added dsn_from option to vary the From: line in DSNs.
2844 PH/18 Flush SMTP output before performing a callout, unless control =
2845 no_callout_flush is set.
2847 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2848 was true (the default) a successful delivery failed to delete the retry
2849 item, thus causing premature timeout of the address. The bug is now
2852 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2854 PH/21 Long custom messages for fakedefer and fakereject are now split up
2855 into multiline responses in the same way that messages for "deny" and
2856 other ACL rejections are.
2858 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2859 with slight modification.
2861 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2862 draft, changing the syntax and factoring some duplicate code".
2864 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2865 for deliveries of the second and subsequent messages over the same SMTP
2868 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2869 ${reduce, with only minor "tidies".
2871 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2873 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2874 expansion side effects.
2876 PH/27 When a message times out after an over-quota error from an Exim-imposed
2877 quota, the bounce message says "mailbox is full". This message was not
2878 being given when it was a system quota that was exceeded. It now should
2881 MH/03 Made $recipients available in local_scan(). local_scan() already has
2882 better access to the recipient list through recipients_list[], but
2883 $recipients can be useful in postmaster-provided expansion strings.
2885 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2886 in the case of a MAIL command with additional options following the
2887 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2888 were accidentally chopped off.
2890 PH/29 SMTP synchronization checks are implemented when a command is read -
2891 there is a check that no more input is waiting when there shouldn't be
2892 any. However, for some commands, a delay in an ACL can mean that it is
2893 some time before the response is written. In this time, more input might
2894 arrive, invalidly. So now there are extra checks after an ACL has run for
2895 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2896 pipelining has not been advertised.
2898 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2900 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2901 returned in the SMTP response when smtp_return_error_details was false.
2902 This has been fixed.
2904 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2905 instead of the C I/O that was originally supplied, because problems were
2906 reported on Solaris.
2908 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2909 Exim which did not show up earlier: it was assuming that a call to
2910 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2911 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2912 was a macro that became an assignment, so it seemed to work. This has
2913 changed to a proper function call with a void return, hence the compile
2914 error. Exim's code has been fixed.
2916 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2919 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2921 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2923 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2924 because it turns out to be too big or there is a timeout) there is no
2925 "Completed" line in the log. When some messages of this type were
2926 selected by exigrep, they were listed as "not completed". Others were
2927 picked up by some special patterns. I have improved the selection
2928 criteria to be more general.
2930 PH/38 The host_find_failed option in the manualroute router can now be set
2931 to "ignore", to completely ignore a host whose IP address cannot be
2932 found. If all hosts are ignored, the behaviour is controlled by the new
2933 host_all_ignored option.
2935 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2936 homing or because of multiple MX records with /mx) generated more than
2937 one IP address, and the following item turned out to be the local host,
2938 all the secondary addresses of the first item were incorrectly removed
2939 from the list, along with the local host and any following hosts (which
2940 is what is supposed to happen).
2942 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2943 whoever called Exim into the -H file. In the case of the daemon it was
2944 behaving confusingly. When first started, it used values for whoever
2945 started the daemon, but after a SIGHUP it used the Exim user (because it
2946 calls itself on a restart). I have changed the code so that it now always
2949 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2950 message are rejected with the same error (e.g. no authentication or bad
2951 sender address), and a DATA command is nevertheless sent (as can happen
2952 with PIPELINING or a stupid MUA), the error message that was given to the
2953 RCPT commands is included in the rejection of the DATA command. This is
2954 intended to be helpful for MUAs that show only the final error to their
2957 PH/42 Another patch from the Sieve maintainer.
2959 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2960 Eximstats - Fixed some broken HTML links and added missing column headers
2962 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2963 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2965 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2967 PH/43 Yet another patch from the Sieve maintainer.
2969 PH/44 I found a way to check for a TCP/IP connection going away before sending
2970 the response to the final '.' that terminates a message, but only in the
2971 case where the client has not sent further data following the '.'
2972 (unfortunately, this is allowed). However, in many cases there won't be
2973 any further data because there won't be any more messages to send. A call
2974 to select() can be used: if it shows that the input is "ready", there is
2975 either input waiting, or the socket has been closed. An attempt to read
2976 the next input character can distinguish the two cases. Previously, Exim
2977 would have sent an OK response which the client would never have see.
2978 This could lead to message repetition. This fix should cure that, at
2979 least in a lot of common cases.
2981 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2982 advertised in response to EHLO.
2988 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2989 fixed by 4.65/MH/01 (is this a record?) are fixed:
2991 (i) An empty string was always treated as zero by the numeric comparison
2992 operators. This behaviour has been restored.
2994 (ii) It is documented that the numeric comparison operators always treat
2995 their arguments as decimal numbers. This was broken in that numbers
2996 starting with 0 were being interpreted as octal.
2998 While fixing these problems I realized that there was another issue that
2999 hadn't been noticed. Values of message_size_limit (both the global option
3000 and the transport option) were treated as octal if they started with 0.
3001 The documentation was vague. These values are now always treated as
3002 decimal, and I will make that clear in the documentation.
3008 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3009 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3012 MH/01 Don't check that the operands of numeric comparison operators are
3013 integers when their expansion is in "skipping" mode (fixes bug
3014 introduced by 4.64-PH/07).
3016 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3017 child addresses, Exim now panics and dies. Previously, because the count
3018 is held in a short int, deliveries were likely to be lost. As such a
3019 large number of recipients for a single message is ridiculous
3020 (performance will be very, very poor), I have chosen to impose a limit
3021 rather than extend the field.
3027 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3028 leftover -K file (the existence of which was triggered by #402).
3029 While we were at it, introduced process PID as part of the -K
3030 filename. This should rule out race conditions when creating
3033 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3034 processing considerably. Previous code took too long for large mails,
3035 triggering a timeout which in turn triggers #401.
3037 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3038 in the DK code in transports.c. sendfile() is not really portable,
3039 hence the _LINUX specificness.
3041 TF/01 In the add_headers option to the mail command in an Exim filter,
3042 there was a bug that Exim would claim a syntax error in any
3043 header after the first one which had an odd number of characters
3046 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3047 callout verification, Exim cached a "reject" for the entire domain. This
3048 is correct for most verifications, but it is not correct for a recipient
3049 verification with use_sender or use_postmaster set, because in that case
3050 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3051 case of MAIL FROM:<> rejection from other early rejections (e.g.
3052 rejection of HELO). When verifying a recipient using a non-null MAIL
3053 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3054 Whatever the result of the callout, the value of the domain cache is
3055 left unchanged (for any other kind of callout, getting as far as trying
3056 RCPT means that the domain itself is ok).
3058 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3061 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3062 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3063 session, the daemon ignores ECONNECT errors and logs others; it now
3064 ignores EPIPE as well.
3066 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3067 (quoted-printable decoding).
3069 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3070 later the small subsequent patch to fix an introduced bug.
3072 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3074 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3076 PH/08 An error is now given if message_size_limit is specified negative.
3078 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3079 to be given (somewhat) arbitrary names.
3081 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3084 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3085 miscellaneous code fixes
3087 PH/10 Added the log_reject_target ACL modifier to specify where to log
3090 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3091 hostname. This is wrong, because it relates to the incoming message (and
3092 probably the interface on which it is arriving) and not to the outgoing
3093 callout (which could be using a different interface). This has been
3094 changed to use the value of the helo_data option from the smtp transport
3095 instead - this is what is used when a message is actually being sent. If
3096 there is no remote transport (possible with a router that sets up host
3097 addresses), $smtp_active_hostname is used.
3099 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3100 tweaks were necessary in order to get it to work (see also 21 below):
3101 (a) The code assumed that strncpy() returns a negative number on buffer
3102 overflow, which isn't the case. Replaced with Exim's string_format()
3104 (b) There were several signed/unsigned issues. I just did the minimum
3105 hacking in of casts. There is scope for a larger refactoring.
3106 (c) The code used strcasecmp() which is not a standard C function.
3107 Replaced with Exim's strcmpic() function.
3108 (d) The code set only $1; it now sets $auth1 as well.
3109 (e) A simple test gave the error "authentication client didn't specify
3110 service in request". It would seem that Dovecot has changed its
3111 interface. Fortunately there's a specification; I followed it and
3112 changed what the client sends and it appears to be working now.
3114 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3117 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3118 address A is aliased to B and C, where B exists and C does not. Without
3119 -v the output is "A verified" because verification stops after a
3120 successful redirection if more than one address is generated. However,
3121 with -v the child addresses are also verified. Exim was outputting "A
3122 failed to verify" and then showing the successful verification for C,
3123 with its parentage. It now outputs "B failed to verify", showing B's
3124 parentage before showing the successful verification of C.
3126 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3127 look up a TXT record in a specific list after matching in a combined
3130 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3131 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3132 they consult the DNS. I had assumed they would set it the way they
3133 wanted; and indeed my experiments on Linux seem to show that in some
3134 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3135 To be on the safe side, however, I have now made the interface to
3136 host_find_byname() similar to host_find_bydns(), with an argument
3137 containing the DNS resolver options. The host_find_byname() function now
3138 sets these options at its start, just as host_find_bydns() does. The smtp
3139 transport options dns_qualify_single and dns_search_parents are passed to
3140 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3141 of host_find_byname() use the default settings of RES_DEFNAMES
3142 (qualify_single) but not RES_DNSRCH (search_parents).
3144 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3145 spool_read_header() do less string testing, by means of a preliminary
3146 switch on the second character of optional "-foo" lines. (This is
3147 overdue, caused by the large number of possibilities that now exist.
3148 Originally there were few.) While I was there, I also converted the
3149 str(n)cmp tests so they don't re-test the leading "-" and the first
3150 character, in the hope this might squeeze out yet more improvement.
3152 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3153 flag allowing group syntax was set by the header_syntax check but not
3154 turned off, possible causing trouble later; (2) The flag was not being
3155 set at all for the header_verify test, causing "group"-style headers to
3156 be rejected. I have now set it in this case, and also caused header_
3157 verify to ignore an empty address taken from a group. While doing this, I
3158 came across some other cases where the code for allowing group syntax
3159 while scanning a header line wasn't quite right (mostly, not resetting
3160 the flag correctly in the right place). These bugs could have caused
3161 trouble for malformed header lines. I hope it is now all correct.
3163 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3164 with the "reply" argument non-NULL. The code, however (which originally
3165 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3166 but it didn't always do it. This confused somebody who was copying the
3167 code for some other use. I have removed all the tests.
3169 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3170 feature that was used to support insecure browsers during the U.S. crypto
3171 embargo. It requires special client support, and Exim is probably the
3172 only MTA that supported it -- and would never use it because real RSA is
3173 always available. This code has been removed, because it had the bad
3174 effect of slowing Exim down by computing (never used) parameters for the
3175 RSA_EXPORT functionality.
3177 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3178 authenticator to fail if there's a tab character in the incoming data
3179 (there should never be unless someone is messing about, as it's supposed
3180 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3181 if the connection is using TLS or if the remote IP is the same as the
3182 local IP, and the "valid-client-cert option" if a client certificate has
3185 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3186 authenticators. This can be used for authorization after authentication
3187 succeeds. (In the case of plaintext, it servers for both authentication
3190 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3191 if any retry times were supplied.
3193 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3194 connection, where there is no client IP address to check. In this
3195 situation, the verify now always succeeds.
3197 PH/25 Applied John Jetmore's -Mset patch.
3199 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3201 PH/27 In a string expansion for a processed (not raw) header when multiple
3202 headers of the same name were present, leading whitespace was being
3203 removed from all of them, but trailing whitespace was being removed only
3204 from the last one. Now trailing whitespace is removed from each header
3205 before concatenation. Completely empty headers in a concatenation (as
3206 before) are ignored.
3208 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3209 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3211 PH/29 [Removed. This was a change that I later backed out, and forgot to
3212 correct the ChangeLog entry (that I had efficiently created) before
3213 committing the later change.]
3215 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3216 address errors (4xx response to RCPT) over the same connection as other
3217 messages routed to the same hosts. Such deliveries are always "forced",
3218 so retry times are not inspected. This resulted in far too many retries
3219 for the affected addresses. The effect occurred only when there were more
3220 hosts than the hosts_max_try setting in the smtp transport when it had
3221 the 4xx errors. Those hosts that it had tried were not added to the list
3222 of hosts for which the message was waiting, so if all were tried, there
3223 was no problem. Two fixes have been applied:
3225 (i) If there are any address or message errors in an SMTP delivery, none
3226 of the hosts (tried or untried) are now added to the list of hosts
3227 for which the message is waiting, so the message should not be a
3228 candidate for sending over the same connection that was used for a
3229 successful delivery of some other message. This seems entirely
3230 reasonable: after all the message is NOT "waiting for some host".
3231 This is so "obvious" that I'm not sure why it wasn't done
3232 previously. Hope I haven't missed anything, but it can't do any
3233 harm, as the worst effect is to miss an optimization.
3235 (ii) If, despite (i), such a delivery is accidentally attempted, the
3236 routing retry time is respected, so at least it doesn't keep
3237 hammering the server.
3239 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3240 in ${readsocket because some servers need this prod.
3242 PH/32 Added some extra debug output when updating a wait-xxx database.
3244 PH/33 The hint "could be header name not terminated by colon", which has been
3245 given for certain expansion errors for a long time, was not being given
3246 for the ${if def:h_colon_omitted{... case.
3248 PH/34 The spec says: "With one important exception, whenever a domain list is
3249 being scanned, $domain contains the subject domain." There was at least
3250 one case where this was not true.
3252 PH/35 The error "getsockname() failed: connection reset by peer" was being
3253 written to the panic log as well as the main log, but it isn't really
3254 panic-worthy as it just means the connection died rather early on. I have
3255 removed the panic log writing for the ECONNRESET error when getsockname()
3258 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3259 runs only) independently of the message's sender address. This meant
3260 that, if the 4xx error was in fact related to the sender, a different
3261 message to the same recipient with a different sender could confuse
3262 things. In particular, this can happen when sending to a greylisting
3263 server, but other circumstances could also provoke similar problems.
3264 I have changed the default so that the retry time for these errors is now
3265 based a combination of the sender and recipient addresses. This change
3266 can be overridden by setting address_retry_include_sender=false in the
3269 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3270 remote server are returned as part of bounce messages. This was not
3271 happening for LMTP over a pipe (the lmtp transport), but now it is the
3272 same for both kinds of LMTP.
3274 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3275 in header lines that were in fact CNAMEs. This is no longer the case.
3277 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3278 and queue runs started by the daemon processed all messages. This has
3279 been fixed so that -R and -S can now usefully be given with -q<time>.
3281 PH/40 Import PCRE release 6.7 (fixes some bugs).
3283 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3285 PH/42 Give an error if -q is specified more than once.
3287 PH/43 Renamed the variables $interface_address and $interface_port as
3288 $received_ip_address and $received_port, to make it clear that these
3289 values apply to message reception, and not to the outgoing interface when
3290 a message is delivered. (The old names remain recognized, of course.)
3292 PH/44 There was no timeout on the connect() call when using a Unix domain
3293 socket in the ${readsocket expansion. There now is.
3295 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3296 be meaningful with "accept".
3298 SC/01 Eximstats V1.43
3299 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3301 SC/02 Eximstats V1.44
3302 Use a glob alias rather than an array ref in the generated
3303 parser. This improves both readability and performance.
3305 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3306 Collect SpamAssassin and rejection statistics.
3307 Don't display local sender or destination tables unless
3308 there is data to show.
3309 Added average volumes into the top table text output.
3311 SC/04 Eximstats V1.46
3312 Collect data on the number of addresses (recipients)
3313 as well as the number of messages.
3315 SC/05 Eximstats V1.47
3316 Added 'Message too big' to the list of mail rejection
3317 reasons (thanks to Marco Gaiarin).
3319 SC/06 Eximstats V1.48
3320 Mainlog lines which have GMT offsets and are too short to
3321 have a flag are now skipped.
3323 SC/07 Eximstats V1.49 (Alain Williams)
3324 Added the -emptyok flag.
3326 SC/08 Eximstats V1.50
3327 Fixes for obtaining the IP address from reject messages.
3329 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3330 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3331 whitespace changes from 4.64-PH/27
3333 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3336 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3337 are found, allow negative numbers in numeric criteria)
3339 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3341 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3342 to match changes made in 4.64-PH/43
3344 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3346 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3347 "domain" is actually an IP address, and force a failure. This locks out
3348 those revolvers/nameservers that support "A-for-A" lookups, in
3349 contravention of the specifications.
3351 PH/48 When a host name was looked up from an IP address, and the subsequent
3352 forward lookup of the name timed out, the host name was left in
3353 $sender_host_name, contrary to the specification.
3355 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3356 restricted to single-key lookups, Exim was not diagnosing an error if
3357 * or *@ was used with a query-style lookup.
3359 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3361 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3362 long ago, but noone interested enough thought of it. Let's just say that
3363 the "1.1" means that there are some new functions that weren't there at
3364 some point in the past.
3366 PH/51 Error processing for expansion failure of helo_data from an smtp
3367 transport during callout processing was broken.
3369 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3370 tested/used via the -bh/-bhc/-bs options.
3372 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3373 bug, fixed in subsequent PCRE releases).
3375 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3376 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3378 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3384 SC/01 Use a glob alias rather than an array ref in eximstats generated
3385 parser. This improves both readability and performance.
3387 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3388 Don't display local sender or destination tables in eximstats unless
3389 there is data to show.
3390 Added average volumes into the eximstats top table text output.
3392 SC/03 Collect data on the number of addresses (recipients) as well
3393 as the number of messages in eximstats.
3395 TF/01 Correct an error in the documentation for the redirect router. Exim
3396 does (usually) call initgroups() when daemonizing.
3398 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3399 with consistent privilege compared to when running as a daemon.
3401 TF/03 Note in the spec that $authenticated_id is not set for local
3402 submissions from trusted users.
3404 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3405 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3407 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3408 by adding some example configuration directives to the default
3409 configuration file. A little bit of work is required to uncomment the
3410 directives and define how usernames and passwords are checked, but
3411 there is now a framework to start from.
3413 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3414 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3415 without this. I don't know how relevant this is to other LDAP libraries.
3417 PH/02 Add the verb name to the "unknown ACL verb" error.
3419 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3421 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3423 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3424 directory not expanded when it should be if an expanded home directory
3425 was set for the address (which is overridden by the transport).
3427 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3430 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3431 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3432 because it is too late at that time, and has no effect.
3434 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3435 security issue with \' (bugzilla #107). I could not use the
3436 PQescapeStringConn() function, because it needs a PGconn value as one of
3439 PH/08 When testing addresses using -bt, indicate those final addresses that
3440 are duplicates that would not cause an additional delivery. At least one
3441 person was confused, thinking that -bt output corresponded to deliveries.
3442 (Suppressing duplicates isn't a good idea as you lose the information
3443 about possibly different redirections that led to the duplicates.)
3445 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3446 systems where poll() doesn't work, in particular OS X.
3448 PH/10 Added more information to debugging output for retry time not reached.
3450 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3451 operations in malware.c.
3453 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3456 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3457 syslog_duplication set false, log lines that would normally be written
3458 both the the main log and to the reject log were not written to syslog at
3461 PH/14 In the default configuration, change the use of "message" in ACL warn
3462 statements to "add_header".
3464 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3465 not followed by a command (e.g. "seen endif").
3467 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3468 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3471 PH/17 Added extra conditions to the default value of delay_warning_condition
3475 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3476 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3477 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3480 The Auto-Submitted: and various List- headers are standardised, whereas I
3481 don't think Precedence: ever was.
3483 PH/18 Refactored debugging code in route_finduser() to show more information,
3484 in particular, the error code if getpwnam() issues one.
3486 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3487 This is apparently needed in addition to the PH/07 change above to avoid
3488 any possible encoding problems.
3490 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3491 but not after initializing Perl.
3493 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3494 output them only if debugging. By default they are written stderr,
3495 apparently, which is not desirable.
3497 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3500 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3503 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3505 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3506 authenticated or an ident call has been made. Suppress the default
3507 values for $authenticated_id and $authenticated_sender (but permit -oMai
3508 and -oMas) when testing with -bh.
3510 PH/24 Re-jigged the order of the tests in the default configuration so that the
3511 tests for valid domains and recipients precede the DNS black list and CSA
3512 tests, on the grounds that those ones are more expensive.
3514 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3515 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3516 This bug exists in every version of Exim that I still have, right back to
3519 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3520 However, an attempt to turn on case-sensitivity in a regex key by
3521 including (?-i) didn't work because the subject string was already
3522 lowercased, and the effects were non-intuitive. It turns out that a
3523 one-line patch can be used to allow (?-i) to work as expected.
3529 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3530 other effects) broke the use of negated acl sub-conditions.
3532 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3535 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3536 "Deny" causes Exim to reject the incoming connection with a 554 error.
3537 Unfortunately, if there is a major crisis, such as a disk failure,
3538 tcp-wrappers gives "deny", whereas what one would like would be some
3539 kind of temporary error. A kludge has been added to help with this.
3540 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3541 554 error is used if errno is still zero or contains ENOENT (which occurs
3542 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3545 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3547 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3548 errors. Otherwise a message that provokes a temporary error (when other
3549 messages do not) can cause a whole host to time out.
3551 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3552 addresses were routed directly to files or pipes from a redirect router.
3553 File deliveries just didn't batch; pipe deliveries might have suffered
3556 PH/06 A failure to get a lock for a hints database would erroneously always say
3557 "Failed to get write lock", even when it was really a read lock.
3559 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3560 of 0600. This has been changed to use the value of the lockfile_mode
3561 option (which defaults to 0600).
3563 PH/08 Applied small patch from the Sieve maintainer.
3565 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3566 folder from quota calculations, a direct delivery into this folder messed
3567 up the contents of the maildirsize file. This was because the regex was
3568 used only to exclude .Trash (or whatever) when the size of the mailbox
3569 was calculated. There was no check that a delivery was happening into an
3570 excluded directory. This bug has been fixed by ignoring all quota
3571 processing for deliveries into excluded directories.
3573 PH/10 Added the maildirfolder_create_regex option to appendfile.
3579 PH/01 The code for finding all the local interface addresses on a FreeBSD
3580 system running IPv6 was broken. This may well have applied to all BSD
3581 systems, as well as to others that have similar system calls. The broken
3582 code found IPv4 interfaces correctly, but gave incorrect values for the
3583 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3584 that it would not match correctly against @[] and not recognize the IPv6
3587 PH/02 The ipliteral router was not recognizing addresses of the form user@
3588 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3590 PH/03 Added disable_ipv6.
3592 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3593 decoded form, because it is most often used to construct To: headers
3594 lines in autoreplies, and the decoded form may well be syntactically
3595 invalid. However, $reply_address has leading white space removed, and all
3596 newlines turned into spaces so that the autoreply transport does not
3599 PH/05 If group was specified without a user on a router, and no group or user
3600 was specified on a transport, the group from the router was ignored.
3602 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3603 for visible compile-time settings that can be used to change these
3604 numbers, for those that want even more. Backwards compatibility with old
3605 spool files has been maintained. However, going back to a previous Exim
3606 release will lost any variables that are in spool files.
3608 PH/07 Two small changes when running in the test harness: increase delay when
3609 passing a TCP/IP connection to a new process, in case the original
3610 process has to generate a bounce, and remove special handling of
3611 127.0.0.2 (sic), which is no longer necessary.
3613 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3614 be the same on different OS.
3616 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3619 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3620 whether --show-vars was specified or not
3622 JJ/02 exipick: Added support for new ACL variable spool format introduced
3625 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3626 syntactically invalid From: or Reply-to: line, and a filter used this to
3627 generate an autoreply, and therefore failed to obtain an address for the
3628 autoreply, Exim could try to deliver to a non-existent relative file
3629 name, causing unrelated and misleading errors. What now happens is that
3630 it logs this as a hard delivery error, but does not attempt to create a
3633 PH/11 The exinext utility has a -C option for testing purposes, but although
3634 the given file was scanned by exinext itself; it wasn't being passed on
3635 when Exim was called.
3637 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3638 an end-of-file indication when reading a command response.
3640 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3641 compiled. In many other places in Exim, IPv6 addresses are always
3642 recognized, so I have changed this. It also means that IPv4 domain
3643 literals of the form [IPV4:n.n.n.n] are now always recognized.
3645 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3646 used if the router is not running as root, for example, when verifying at
3647 ACL time, or when using -bh. The debugging output from this situation was
3648 non-existent - all you got was a failure to exec. I have made two
3651 (a) Failures to set uid/gid, the current directory, or a process leader
3652 in a subprocess such as that created by queryprogram now generate
3653 suitable debugging output when -d is set.
3655 (b) The queryprogram router detects when it is not running as root,
3656 outputs suitable debugging information if -d is set, and then runs
3657 the subprocess without attempting to change uid/gid.
3659 PH/15 Minor change to Makefile for building test_host (undocumented testing
3662 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3663 additional section of a DNS packet that returns MX or SRV records.
3664 Instead, it always explicitly searches for A/AAAA records. This avoids
3665 major problems that occur when a DNS server includes only records of one
3666 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3667 fixed another bug: if SRV records were looked up and the corresponding
3668 address records were *not* found in the additional section, the port
3669 values from the SRV records were lost.
3671 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3672 using the correct key (the original address) when searching the retry
3673 rules in order to find which one to use for generating the retry hint.
3675 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3676 adding the default one. Similarly, if it contains a Reply-To: header, the
3677 errors_reply_to option, if set, is not used.
3679 PH/19 When calculating a retry time, Exim used to measure the "time since
3680 failure" by looking at the "first failed" field in the retry record. Now
3681 it does not use this if it is later than than the arrival time of the
3682 message. Instead it uses the arrival time. This makes for better
3683 behaviour in cases where some deliveries succeed, thus re-setting the
3684 "first failed" field. An example is a quota failure for a huge message
3685 when small messages continue to be delivered. Without this change, the
3686 "time since failure" will always be short, possible causing more frequent
3687 delivery attempts for the huge message than are intended.
3688 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3690 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3691 $1, $2, $3) because the numerical variables can be reset during some
3692 expansion items (e.g. "match"), thereby losing the authentication data.
3694 PH/21 Make -bV show the size of off_t variables so that the test suite can
3695 decide whether to run tests for quotas > 2G.
3697 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3698 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3699 filecount value is greater than 2G or if a quota value is greater than 2G
3700 on a system where the size of off_t is not greater than 4, a panic error
3703 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3704 never match. The debug and -bh output now contains an explicit error
3705 message indicating a malformed IPv4 address or mask.
3707 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3708 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3709 PH/23 above applies.
3711 PH/25 Do not write to syslog when running in the test harness. The only
3712 occasion when this arises is a failure to open the main or panic logs
3713 (for which there is an explicit test).
3715 PH/26 Added the /no_tell option to "control=freeze".
3717 PH/27 If a host name lookup failed very early in a connection, for example, if
3718 the IP address matched host_lookup and the reverse lookup yielded a name
3719 that did not have a forward lookup, an error message of the form "no IP
3720 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3721 could be logged. Now it outputs the IP address instead of "NULL".
3723 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3724 allows the sender and the authenticated sender to be set when
3725 submitting a message from within Exim. Since child_open_exim() is
3726 documented for local_scan(), the new function should be too.
3728 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3729 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3730 results in an empty string is now treated as unset.
3732 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3734 PH/31 Added sender_verify_fail logging option.
3736 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3737 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3738 tidied the source and removed it altogether.
3740 PH/33 When a queue run was abandoned because the load average was too high, a
3741 log line was always written; now it is written only if the queue_run log
3742 selector is set. In addition, the log line for abandonment now contains
3743 information about the queue run such as the pid. This is always present
3744 in "start" and "stop" lines but was omitted from the "abandon" line.
3746 PH/34 Omit spaces between a header name and the colon in the error message that
3747 is given when verify = headers_syntax fails (if there are lots of them,
3748 the message gets confusing).
3750 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3751 names, as there are now some PTR records that contain slashes. This check
3752 is only to protect against broken name servers that fall over on strange
3753 characters, so the fact that it applies to all lookups doesn't matter.
3755 PH/36 Now that the new test suite is complete, we can remove some of the
3756 special code in Exim that was needed for the old test suite. For example,
3757 sorting DNS records because real resolvers return them in an arbitrary
3758 order. The new test suite's fake resolver always returns records in the
3761 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3762 bounces) except when queue_only is set, to avoid logging races between
3763 the different processes.
3765 PH/38 Panic-die if .include specifies a non-absolute path.
3767 PH/39 A tweak to the "H" retry rule from its user.
3769 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3770 a label. They prevented compilation on older perls.
3772 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3773 a warning to be raised on newish perls.
3775 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3776 on queue. Changes to match documented behaviour of showing count of
3777 messages matching specified criteria.
3779 PH/40 Changed the default ident timeout from 30s to 5s.
3781 PH/41 Added support for the use of login_cap features, on those BSD systems
3782 that have them, for controlling the resources used by pipe deliveries.
3784 PH/42 The content-scanning code uses fopen() to create files in which to put
3785 message data. Previously it was not paying any attention to the mode of
3786 the files. Exim runs with umask(0) because the rest of the code creates
3787 files with open(), and sets the required mode explicitly. Thus, these
3788 files were ending up world-writeable. This was not a big issue, because,
3789 being within the spool directory, they were not world-accessible. I have
3790 created a function called modefopen, which takes an additional mode
3791 argument. It sets umask(777), creates the file, chmods it to the required
3792 mode, then resets the umask. All the relevant calls to fopen() in the
3793 content scanning code have been changed to use this function.
3795 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3796 to 24 hours. This avoids potential overflow problems when processing G
3797 and H retry rules. I suspect nobody ever tinkers with this value.
3799 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3801 PH/45 When the plaintext authenticator is running as a client, the server's
3802 challenges are checked to ensure they are valid base64 strings. By
3803 default, the authentication attempt is cancelled if an invalid string is
3804 received. Setting client_ignore_invalid_base64 true ignores these errors.
3805 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3806 they are received. Thus, the responses can be made to depend on the
3807 challenges. If an invalid string is ignored, an empty string is placed in
3810 PH/46 Messages that are created by the autoreply transport now contains a
3811 References: header, in accordance with RFCs 2822 and 3834.
3813 PH/47 Added authenticated_sender_force to the smtp transport.
3815 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3817 PH/49 Installed latest patch from the Sieve maintainer.
3819 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3820 was also set, the appendfile transport was unnecessarily scanning a
3821 directory of message files (e.g. for maildir delivery) to find the count
3822 of files (along with the size), even though it did not need this
3823 information. It now does the scan only if it needs to find either the
3824 size of the count of files.
3826 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3828 PH/52 Two bugs concerned with error handling when the smtp transport is
3831 (i) Exim was not creating retry information for temporary errors given
3832 for individual recipients after the DATA command when the smtp transport
3833 was used in LMTP mode. This meant that they could be retried too
3834 frequently, and not timed out correctly.
3836 (ii) Exim was setting the flag that allows error details to be returned
3837 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3838 recipients that were returned after the DATA command.
3840 PH/53 This is related to PH/52, but is more general: for any failing address,
3841 when detailed error information was permitted to be returned to the
3842 sender, but the error was temporary, then after the final timeout, only
3843 "retry timeout exceeded" was returned. Now it returns the full error as
3844 well as "retry timeout exceeded".
3846 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3847 do this, and (what is worse) MTAs that accept it.
3849 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3850 will now be deprecated.
3852 PH/56 New os.c-cygwin from the Cygwin maintainer.
3854 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3855 formats (previously only available in exim formats via -bpr, -bpru,
3856 and -bpra. Now also available in native and exiqgrep formats)
3858 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3859 with very large, slow to parse queues
3861 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3863 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3865 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3866 responses to authentication challenges, though it was showing the
3867 challenges; (ii) I've removed the CR characters from the debug output for
3870 PH/58 Allow for the insertion of a newline as well as a space when a string
3871 is turned into more than one encoded-word during RFC 2047 encoding. The
3872 Sieve code now uses this.
3874 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3875 data_4xx, lost_connection, tls_required.
3877 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3878 message was being sent as an SMTP response.
3880 PH/61 Add -l and -k options to exicyclog.
3882 PH/62 When verifying, if an address was redirected to one new address, so that
3883 verification continued, and the new address failed or deferred after
3884 having set something in $address_data, the value of $address_data was not
3885 passed back to the ACL. This was different to the case when no
3886 redirection occurred. The value is now passed back in both cases.
3888 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3889 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3890 use login_cap.h, so on its own it isn't the distinguishing feature. The
3891 new name refers directly to the setclassresources() function.
3893 PH/65 Added configuration files for NetBSD3.
3895 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3897 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3898 is preferred over IPv4.
3900 PH/68 The bounce_return_message and bounce_return_body options were not being
3901 honoured for bounces generated during the reception of non-SMTP messages.
3902 In particular, this applied to messages rejected by the ACL. This bug has
3903 been fixed. However, if bounce_return_message is true and bounce_return_
3904 body is false, the headers that are returned for a non-SMTP message
3905 include only those that have been read before the error was detected.
3906 (In the case of an ACL rejection, they have all been read.)
3908 PH/69 The HTML version of the specification is now built in a directory called
3909 spec_html instead of spec.html, because the latter looks like a path with
3910 a MIME-type, and this confuses some software.
3912 PH/70 Catch two compiler warnings in sieve.c.
3914 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3915 function verify_get_ident() calls ip_connect() to connect a socket, but
3916 if the "connect()" function timed out, ip_connect() used to close the
3917 socket. However, verify_get_ident() also closes the socket later, and in
3918 between Exim writes to the log, which may get opened at this point. When
3919 the socket was closed in ip_connect(), the log could get the same file
3920 descriptor number as the socket. This naturally causes chaos. The fix is
3921 not to close the socket in ip_connect(); the socket should be closed by
3922 the function that creates it. There was only one place in the code where
3923 this was missing, in the iplookup router, which I don't think anybody now
3924 uses, but I've fixed it anyway.
3926 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3927 well as to direct DNS lookups. Otherwise the handling of names in host
3928 lists is inconsistent and therefore confusing.
3934 PH/01 Two changes to the default runtime configuration:
3936 (1) Move the checks for relay_from_hosts and authenticated clients from
3937 after to before the (commented out) DNS black list checks.
3939 (2) Add control=submission to the relay_from_hosts and authenticated
3940 clients checks, on the grounds that messages accepted by these
3941 statements are most likely to be submissions.
3943 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3945 (1) Generate an error if the third argument for the ${prvs expansion is
3948 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3951 (3) Reset the variables that are obtained from the first argument of
3952 ${prvscheck and used in the second argument before leaving the code,
3953 because their memory is reclaimed, so using them afterwards may do
3956 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3957 one (it's much easier than Tom thought :-).
3959 (5) Because of (4), we can now allow for the use of $prvscheck_result
3960 inside the third argument.
3962 PH/03 For some reason, the default setting of PATH when running a command from
3963 a pipe transport was just "/usr/bin". I have changed it to
3966 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3967 anything to be listed in the output from -bV.
3969 PH/05 When a filter generated an autoreply, the entire To: header line was
3970 quoted in the delivery log line, like this:
3972 => >A.N.Other <ano@some.domain> <original@ddress> ...
3974 This has been changed so that it extracts the operative address. There
3975 may be more than one such address. If so, they are comma-separated, like
3978 => >ano@some.domain,ona@other.domain <original@ddress> ...
3980 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3981 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3982 not being looked up in the rDNS to get a host name, Exim was showing the
3983 IP address twice in Received: lines, even though the IP addresses were
3984 identical. For example:
3986 Received: from [1.2.3.4] (helo=[1.2.3.4])
3988 However, if the real host name was known, it was omitting the HELO data
3989 if it matched the actual IP address. This has been tidied up so that it
3990 doesn't show the same IP address twice.
3992 PH/07 When both +timestamp and +memory debugging was on, the value given by
3993 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3994 called by the debug printing, thereby overwriting the timestamp buffer.
3995 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3997 PH/08 When the original message was included in an autoreply transport, it
3998 always said "this is a copy of the message, including all the headers",
3999 even if body_only or headers_only was set. It now gives an appropriate
4002 PH/09 Applied a patch from the Sieve maintainer which:
4004 o fixes some comments
4005 o adds the (disabled) notify extension core
4006 o adds some debug output for the result of if/elsif tests
4007 o points to the current vacation draft in the documentation
4008 and documents the missing references header update
4012 o fixes a bug in processing the envelope test (when testing
4013 multiple envelope elements, the last element determined the
4016 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4017 Electronic Mail") by including:
4019 Auto-submitted: auto-generated
4021 in the messages that it generates (bounce messages and others, such as
4022 warnings). In the case of bounce messages for non-SMTP messages, there was
4023 also a typo: it was using "Auto_submitted" (underscore instead of
4024 hyphen). Since every message generated by Exim is necessarily in response
4025 to another message, thes have all been changed to:
4027 Auto-Submitted: auto-replied
4029 in accordance with these statements in the RFC:
4031 The auto-replied keyword:
4033 - SHOULD be used on messages sent in direct response to another
4034 message by an automatic process,
4036 - MUST NOT be used on manually-generated messages,
4038 - MAY be used on Delivery Status Notifications (DSNs) and Message
4039 Disposition Notifications (MDNs),
4041 - MUST NOT be used on messages generated by automatic or periodic
4042 processes, except for messages which are automatic responses to
4045 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4046 to the default Received: header definition.
4048 PH/12 Added log selector acl_warn_skipped (default on).
4050 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4051 variables because (a) they are in the wrong storage pool and (b) even if
4052 they were copied, it wouldn't work properly because of the caching.
4054 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4055 checking when decoding. Apparently there are clients that generate
4056 overlong encoded strings. Why am I not surprised?
4058 PH/15 If the first argument of "${if match_address" was not empty, but did not
4059 contain an "@" character, Exim crashed. Now it writes a panic log message
4060 and treats the condition as false.
4062 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4064 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4065 extension "envelope-auth". The code is finished and in agreement with
4066 other implementations, but there is no documentation so far and in fact,
4067 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4068 not changing the active code.
4070 Print executed "if" and "elsif" statements when debugging is used. This
4071 helps a great deal to understand what a filter does.
4073 Document more things not specified clearly in RFC3028. I had all this
4074 sorted out, when out of a sudden new issues came to my mind. Oops."
4076 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4079 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4080 canonical form (as documented). However, after a host name lookup from
4081 the IP address, check_host() was doing a simple string comparison with
4082 addresses acquired from the DNS when checking that the found name did
4083 have the original IP as one of its addresses. Since any found IPv6
4084 addresses are likely to be in abbreviated form, the comparison could
4085 fail. Luckily, there already exists a function for doing the comparison
4086 by converting both addresses to binary, so now that is used instead of
4087 the text comparison.
4089 PH/20 There was another similar case to PH/19, when a complete host name was
4090 given in a host list; looking up its IP address could give an abbreviated
4091 form, whereas the current host's name might or might not be abbreviated.
4092 The same fix has been applied.
4098 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4099 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4102 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4103 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4105 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4107 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4108 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4109 Also, for "Auto-Submitted", treat anything other than "no" as
4110 non-personal, in accordance with RFC 3834. (Previously it treated
4111 anything starting "auto-" as non-personal.)
4113 TF/01 The control=submission/name=... option had a problem with syntax
4114 errors if the name included a slash character. The /name= option
4115 now slurps the rest of the string, so it can include any characters
4116 but it must come last in the list of options (after /sender_retain
4119 PH/05 Some modifications to the interface to the fake nameserver for the new
4127 TK/01 Added the "success_on_redirect" address verification option. See
4128 NewStuff for rationale and an example.
4130 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4132 PH/02 Patch to exigrep to allow it to work on syslog lines.
4134 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4135 fread() to skip over the body file's header line, because in Cygwin the
4136 header line is locked and is inaccessible.
4138 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4139 co-exist for some time) to make it clear that it is the Exim ID that is
4140 referenced, not the Message-ID: header line.
4142 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4143 string_format() function, because snprintf() does not exist on all
4146 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4147 ${stat: expansion item.
4149 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4150 protocol synchronization error", to keep the pedants happy.
4152 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4153 well as for IRIX systems, when gcc is being used. See the host.c source
4156 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4158 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4161 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4162 transport and to the smtp transport in LMTP mode.
4164 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4166 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4167 run a filter in a subprocess. This could lead to confusion in subsequent
4168 lookups in the parent process. There should also be a search_tidyup() at
4169 the end of the subprocess.
4171 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4172 only if the host matched helo_try_verify_hosts, which caused the
4173 verification to occur when the EHLO/HELO command was issued. The ACL just
4174 tested the remembered result. Now, if a previous verification attempt has
4175 not happened, "verify = helo" does it there and then.
4177 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4179 TK/03 Fix log output including CR from clamd.
4181 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4182 exist provoked a memory error which could cause a segfault.
4184 PH/15 Installed PCRE 6.2
4186 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4188 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4189 of the problem. Specifically, suggested +O2 rather than +O1 for the
4192 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4194 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4195 record, the port was not correctly logged when the outgoing_port log
4196 selector was set (it logged the transort's default port).
4198 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4199 fallback_hosts, and "hosts" in the smtp transport.
4201 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4202 host errors such as "Connection refused".
4204 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4205 authentication with radiusclient 0.4.9:
4207 - Error returned from rc_read_config was caught wrongly
4208 - Username/password not passed on to radius server due to wrong length.
4210 The presumption is that some radiusclient API changes for 4.51/PH/17
4211 were not taken care of correctly. The code is still untested by me (my
4212 Linux distribution still has 0.3.2 of radiusclient), but it was
4213 contributed by a Radius user.
4215 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4216 expanding the "port" option of the smtp transport.
4218 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4219 while reading a MIME header. Thanks to Tom Hughes for a patch.
4221 PH/24 Include config.h inside local_scan.h so that configuration settings are
4224 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4225 that in an ACL for RCPT (for example), you can examine exactly what was
4228 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4229 commands, but it was not correctly comparing the address with the actual
4230 client host address. Thus, it would show the EHLO address in Received:
4231 header lines when this was not necessary.
4233 PH/27 Added the % operator to ${eval:}.
4235 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4236 it should be ignoring failures (because with -C, for example, it has lost
4237 privilege). It wasn't ignoring creation failures other than "already
4240 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4243 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4244 because an input line was too long, either on its own, or by virtue of
4245 too many continuations, the temporary file was not being removed, and the
4246 return code was incorrect.
4248 PH/31 Missing "BOOL" in function definition in filtertest.c.
4250 PH/32 Applied Sieve patches from the maintainer.
4252 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4254 PH/33 Added "verify = not_blind".
4256 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4257 Local/Makefile (with some defaults set). These are used in built scripts
4258 such as exicyclog, but they have never been used in the exim_install
4259 script (though there are many overriding facilities there). I have
4260 arranged that the exim_install script now takes note of these two
4263 PH/35 Installed configuration files for Dragonfly.
4265 PH/36 When a locally submitted message by a trusted user did not contain a
4266 From: header, and the sender address was obtained from -f or from an SMTP
4267 MAIL command, and the trusted user did not use -F to supply a sender
4268 name, $originator_name was incorrectly used when constructing a From:
4269 header. Furthermore, $originator_name was used for submission mode
4270 messages from external hosts without From: headers in a similar way,
4271 which is clearly wrong.
4273 PH/37 Added control=suppress_local_fixups.
4275 PH/38 When log_selector = +received_sender was set, and the addition of the
4276 sender made the log line's construction buffer exactly full, or one byte
4277 less than full, an overflow happened when the terminating "\n" was
4280 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4281 when the result of a list match is failure because a DNS lookup failed.
4283 PH/40 RM_COMMAND is now used in the building process.
4285 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4286 the "build-* directories that it finds.
4288 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4289 address in a domain literal was a prefix of an interface address.
4291 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4292 when verifying a sender address, unless rewrite_headers is false.
4294 PH/44 Wrote a long comment about why errors_to addresses are verified as
4295 recipients, not senders.
4297 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4298 the ratelimit ACL was added.
4300 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4302 PH/46 Added extra information about PostgreSQL errors to the error string.
4304 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4305 avoiding the need to install special zones in a real server. This is
4306 backwards compatible; if it can't find the fake resolver, it drops back.
4307 Thus, both old and new test suites can be run.
4309 TF/02 Added util/ratelimit.pl
4311 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4312 clock is set back in time.
4314 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4315 Candler <B.Candler@pobox.com>.
4317 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4318 OK for addresses that are the result of redirections.
4320 PH/48 A number of further additions for the benefit of the new test suite,
4321 including a fake gethostbyname() that interfaces to the fake DNS resolver
4324 TF/06 The fix for widen_domains has also been applied to qualify_single and
4325 search_parents which are the other dnslookup options that can cause
4328 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4331 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4333 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4334 service" if it encountered an MX record with an empty target hostname.
4335 The message is now "an MX or SRV record indicated no SMTP service".
4337 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4338 if the DNS of the sending site is misconfigured. This is quite a
4339 common situation. This change restores the behaviour of treating a
4340 helo verification defer as a failure.
4342 PH/51 If self=fail was set on a router, the bounce message did not include the
4343 actual error message.
4349 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4351 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4352 command itself ended in error, the underlying message about the transport
4353 filter timeout was being overwritten with the pipe command error. Now the
4354 underlying error message should be appended to the second error message.
4356 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4358 PH/02 Reduce the amount of output that "make" produces by default. Full output
4359 can still be requested.
4361 PH/03 The warning log line about a condition test deferring for a "warn" verb
4362 was being output only once per connection, rather than after each
4363 occurrence (because it was using the same function as for successful
4364 "warn" verbs). This seems wrong, so I have changed it.
4366 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4367 it should not have, which might have caused a crash in the right
4368 circumstances, but probably never did.
4370 PH/04 Installed a modified version of Tony Finch's patch to make submission
4371 mode fix the return path as well as the Sender: header line, and to
4372 add a /name= option so that you can make the user's friendly name appear
4375 TF/03 Added the control = fakedefer ACL modifier.
4377 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4378 Mark Lowes for thorough testing.
4380 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4382 TK/03 Merged latest SRS patch from Miles Wilton.
4384 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4385 with the definition in sysexits.h (which is #included earlier).
4386 Fortunately, Exim does not actually use EX_OK. The code used to try to
4387 preserve the sysexits.h value, by assuming that macro definitions were
4388 scanned for macro replacements. I have been disabused of this notion,
4389 so now the code just undefines EX_OK before #including unistd.h.
4391 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4392 in the smtp transport. When a block could not be written in a single
4393 write() function, the timeout was being re-applied to each part-write.
4394 This seems wrong - if the receiver was accepting one byte at a time it
4395 would take for ever. The timeout is now adjusted when this happens. It
4396 doesn't have to be particularly precise.
4398 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4399 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4401 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4402 without a domain if the check to <postmaster@domain> fails.
4404 SC/01 Eximstats: added -xls and the ability to specify output files
4405 (patch written by Frank Heydlauf).
4407 SC/02 Eximstats: use FileHandles for outputting results.
4409 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4411 SC/04 Eximstats: fixed display of large numbers with -nvr option
4413 SC/05 Eximstats: fixed merging of reports with empty tables.
4415 SC/06 Eximstats: added the -include_original_destination flag
4417 SC/07 Eximstats: removed tabs and trailing whitespace.
4419 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4421 TK/06 MBOX spool code: Add real "From " MBOX separator line
4422 so the .eml file is really in mbox format (even though
4423 most programs do not really care). Patch from Alex Miller.
4425 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4426 The latter is generated from $received_to and is only set if the
4427 message has one envelope recipient. SA can use these headers,
4428 obviously out-of-the-box. Patch from Alex Miller.
4430 PH/08 The ${def test on a variable was returning false if the variable's
4431 value was "0", contrary to what the specification has always said!
4432 The result should be true unless the variable is empty.
4434 PH/09 The syntax error of a character other than { following "${if
4435 def:variable_name" (after optional whitespace) was not being diagnosed.
4436 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4437 accidental colon was present, for example, could give incorrect results.
4439 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4440 result is used (not including appendfile, where other changes are about
4443 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4444 This involved changing a lot of size variables from int to off_t. It
4445 should work with maildirs and everything.
4447 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4448 spamd dying while we are connected to it.
4450 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4453 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4454 function for BDB 4.3.
4456 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4458 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4459 the sources for PCRE have been re-arranged and more files are now
4462 PH/15 The code I had for printing potentially long long variables in PH/11
4463 above was not the best (it lost precision). The length of off_t variables
4464 is now inspected at build time, and an appropriate printing format (%ld
4465 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4466 to be "long long int" or "long int". This is needed for the internal
4467 formatting function string_vformat().
4469 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4470 the configuration file to be ":syslog", then the script "guesses" where
4471 the logs files are, rather than using the compiled in default. In our
4472 case the guess is not the same as the compiled default, so the script
4473 suddenly stopped working when I started to use syslog. The patch checks
4474 to see if log_file_path is "". If so, it attempts to read it from exim
4475 with no configuration file to get the compiled in version, before it
4476 falls back to the previous guessing code."
4478 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4479 implementing BATV in an Exim configuration. See NewStuff for the gory
4482 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4483 Makefile that are specific to HP-UX.
4485 PH/18 If the "use_postmaster" option was set for a recipient callout together
4486 with the "random" option, the postmaster address was used as the MAIL
4487 FROM address for the random test, but not for the subsequent recipient
4488 test. It is now used for both.
4490 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4491 patch removes a few documentation additions to RFC 3028, because the
4492 latest draft now contains them. It adds the new en;ascii-case comparator
4493 and a new error check for 8bit text in MIME parts. Comparator and
4494 require names are now matched exactly. I enabled the subaddress
4495 extension, but it is not well tested yet (read: it works for me)."
4497 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4498 rework some of the code of TK/09 above to avoid the hardwired use of
4499 "%lld" and "long long". Replaced the call to snprintf() with a call to
4502 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4503 records point to non-existent hosts", "retry timeout exceeded", and
4504 "retry time not reached for any host after a long failure period".
4506 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4507 experimental DomainKeys support:
4509 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4510 (2) On an error such as an illegally used "control", the wrong name for
4511 the control was given.
4513 These problems did NOT occur unless DomainKeys support was compiled.
4515 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4517 PH/24 Added ${if match_ip condition.
4519 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4520 databases so that it will be absolutely obvious if a crash occurs in the
4521 DB library. This is a regular occurrence (often caused by mis-matched
4524 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4525 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4526 picked up on a user's system that detects such things. There doesn't seem
4527 to be a gcc warning option for this - only an attribute that has to be
4528 put on the function's prototype. It seems that in Fedora Core 4 they have
4529 set this on a number of new functions. No doubt there will be more in due
4532 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4533 specify a transport. However, if an address that was verified by such a
4534 router was the subject of a callout, Exim crashed because it tried to
4535 read the rcpt_include_affixes from the non-existent transport. Now it
4536 just assumes that the setting of that option is false. This bug was
4537 introduced by 4.51/PH/31.
4539 PH/28 Changed -d+all to exclude +memory, because that information is very
4540 rarely of interest, but it makes the output a lot bigger. People tend to
4541 do -d+all out of habit.
4543 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4544 code in os-type was giving problems when libc.so lives in lib64, like on
4547 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4548 aren't the modern standard, and it seems that some systems' include files
4549 don't always have them. Exim was already checking for some of the newer
4550 ones like T_AAAA, and defining it itself. I've added checks for all the
4551 record types that Exim uses.
4553 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4554 not automatically generating a new one, as it is supposed to. This
4555 prevented TLS from working. If the file did exist, but contained invalid
4556 data, a new version was generated, as expected. It was only the case of a
4557 non-existent file that was broken.
4559 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4560 with a change in libdomainkeys > 0.64.
4562 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4563 from DNS. If the selector record carries the flag, it now has
4564 precedence over the domain-wide flag.
4566 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4568 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4569 the use of a port name that isn't defined in /etc/services) occurred, the
4570 message was deferred as in a normal delivery, and thus remained on the
4571 spool, instead of being failed because of the mua_wrapper setting. This
4572 is now fixed, and I tidied up some of the mua_wrapper messages at the
4575 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4576 the messages in a hash of arrays rather than using individual hashes.
4577 This is a bit cleaner and results in dramatic memory savings, albeit
4578 at a slight CPU cost.
4580 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4581 as requested by Marc Sherman.
4583 SC/10 Eximstats: added histograms for user specified patterns as requested
4586 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4588 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4589 fopen() in the content-scanning modules that did not already have it.
4595 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4596 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4598 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4600 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4602 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4603 to test Sieve filters that use "vacation".
4605 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4606 that changes the way the GnuTLS parameters are stored in the cache file.
4607 The new format can be generated externally. For backward compatibility,
4608 if the data in the cache doesn't make sense, Exim assumes it has read an
4609 old-format file, and it generates new data and writes a new file. This
4610 means that you can't go back to an older release without removing the
4613 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4614 work if there are any delivery delays because "one_time" forces the
4615 parent to be marked "delivered", so its unseen clone is never tried
4616 again. For this reason, Exim now forbids the simultaneous setting of
4619 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4620 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4621 entry if you want to know the details. The fix, however, neglected to
4622 consider the case when local delivery batching is involved. The test for
4623 "previously delivered" was not happening when checking to see if an
4624 address could be batched with a previous (undelivered) one; under
4625 certain circumstances this could lead to multiple deliveries to the same
4628 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4629 in its include files, and this causes problems building Exim.
4631 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4632 header_syntax) but Exim was just ignoring anything given after a slash.
4633 In particular, this caused confusion with an attempt to use "verify =
4634 reverse_host_lookup/defer_ok". An error is now given when options are
4635 supplied for verify items that do not have them. (Maybe reverse_host_
4636 lookup should have a defer_ok option, but that's a different point.)
4638 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4639 defined by RFC 821) to 2048, because there were problems with some AUTH
4640 commands, and RFC 1869 says the size should be increased for extended
4641 SMTP commands that take arguments.
4643 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4646 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4647 "unknown" error; now it says that the functionality isn't in the binary.
4649 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4650 an address' error message when a string expansion fails (syntax or
4651 whatever). Otherwise the password may appear in the log. Following change
4652 PH/42 below, there is no longer a chance of it appearing in a bounce
4655 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4657 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4658 of its addresses was ever tried. (Bugzilla bug #2.)
4660 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4661 the result incorrectly in the debug output. (It correctly added a newline
4662 to what was transported.)
4664 TF/01 Added $received_time.
4666 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4667 commented out examples of how to interface to a virus scanner and to
4668 SpamAssassin. Also added commented examples of av_scanner and
4669 spamd_address settings.
4671 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4672 and controls are allowed in which ACLs. There were a couple of minor
4673 errors. Some of the entries in the conditions table (which is a table of
4674 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4675 negation of where the condition IS allowed.
4677 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4679 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4680 header file does not have a version number, so I've had to invent a new
4681 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4682 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4683 radiusclient), but it was contributed by a Radius user.
4685 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4686 files or directories, for OpenSSL.
4688 PH/19 When an Exim process that is running as root has to create an Exim log
4689 file, it does so in a subprocess that runs as exim:exim so as to get the
4690 ownership right at creation (otherwise, other Exim processes might see
4691 the file with the wrong ownership). There was no test for failure of this
4692 fork() call, which would lead to the process getting stuck as it waited
4693 for a non-existent subprocess. Forks do occasionally fail when resources
4694 run out. I reviewed all the other calls to fork(); they all seem to check
4697 PH/20 When checking for unexpected SMTP input at connect time (before writing
4698 the banner), Exim was not dealing correctly with a non-positive return
4699 from the read() function. If the client had disconnected by this time,
4700 the result was a log entry for a synchronization error with an empty
4701 string after "input=" when read() returned zero. If read() returned -1
4702 (an event I could not check), uninitialized data bytes were printed.
4703 There were reports of junk text (parts of files, etc) appearing after
4706 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4708 PH/22 Added support for macro redefinition, and (re)definition in between
4709 driver and ACL definitions.
4711 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4712 forgetting to use the resulting value; it was using the unexpanded value.
4714 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4715 hadn't been configured. The fix is from Juergen Kreileder, who
4716 understands it better than I do:
4718 "Here's what I see happening with three configured cyrus_sasl
4719 authenticators configured (plain, login, cram-md5):
4721 On startup auth_cyrus_sasl_init() gets called for each of these.
4722 This means three calls to sasl_listmech() without a specified mech_list.
4723 => SASL tests which mechs of all available mechs actually work
4724 => three warnings about OTP not working
4725 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4727 With the patch, sasl_listmech() also gets called three times. But now
4728 SASL's mech_list option is set to the server_mech specified in the the
4729 authenticator. Or in other words, the answer from sasl_listmech()
4730 gets limited to just the mech you're testing for (which is different
4732 => the return list contains just 'plain' or 'login', 'cram-md5' or
4733 nothing depending on the value of ob->server_mech.
4735 I've just tested the patch: Authentication still works fine,
4736 unavailable mechs specified in the exim configuration are still
4737 caught, and the auth.log warnings about OTP are gone."
4739 PH/25 When debugging is enabled, the contents of the command line are added
4740 to the debugging output, even when log_selector=+arguments is not
4743 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4744 answer is "GNU", and only if the return is "GNU/something" is the answer
4747 PH/27 $acl_verify_message is now set immediately after the failure of a
4748 verification in an ACL, and so is available in subsequent modifiers. In
4749 particular, the message can be preserved by coding like this:
4751 warn !verify = sender
4752 set acl_m0 = $acl_verify_message
4754 Previously, $acl_verify_message was set only while expanding "message"
4755 and "log_message" when a very denied access.
4757 PH/28 Modified OS/os.c-Linux with
4759 -#ifndef OS_LOAD_AVERAGE
4760 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4762 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4763 nomenclature these days.)
4765 PH/29 Installed patch from the Sieve maintainer that adds the options
4766 sieve_useraddress and sieve_subaddress to the redirect router.
4768 PH/30 In these circumstances:
4769 . Two addresses routed to the same list of hosts;
4770 . First host does not offer TLS;
4771 . First host accepts first address;
4772 . First host gives temporary error to second address;
4773 . Second host offers TLS and a TLS session is established;
4774 . Second host accepts second address.
4775 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4776 and peerdn, if requested) that were in fact used only for the second
4779 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4780 attention to any local part prefix or suffix that was matched by the
4781 router that accepted the address. It now behaves in the same way as it
4782 does for delivery: the affixes are removed from the local part unless
4783 rcpt_include_affixes is set on the transport.
4785 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4786 timeout during the DATA phase of an incoming message.
4788 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4789 applied a patch sanctioned by the Sieve maintainer.
4791 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4792 the uid or gid is negative. A case of a negative gid caused this to be
4793 noticed. The fix allows for either to be negative.
4795 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4796 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4799 PH/36 The change PH/12 above was broken. Fixed it.
4801 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4802 the grounds that routing the same address twice would always produce the
4803 same answer. This might have been true once, but it is certainly no
4804 longer true now. Routing a child address may depend on the previous
4805 routing that produced that child. Some complicated redirection strategies
4806 went wrong when messages had multiple recipients, and made Exim's
4807 behaviour dependent on the order in which the addresses were given.
4809 I have moved the duplicate checking until after the routing is complete.
4810 Exim scans the addresses that are assigned to local and remote
4811 transports, and removes any duplicates. This means that more work will be
4812 done, as duplicates will always all be routed, but duplicates are
4813 presumably rare, so I don't expect this is of any significance.
4815 For deliveries to pipes, files, and autoreplies, the duplicate checking
4816 still happens during the routing process, since they are not going to be
4819 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4820 It corrects a timeout issue with spamd. This is Ian's comment: "The
4821 background is that sometimes spamd either never reads data from a
4822 connection it has accepted, or it never writes response data. The exiscan
4823 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4824 blindly assumes that writes won't block so it may never time out."
4826 PH/39 Allow G after quota size as well as K and M.
4828 PH/40 The value set for $authenticated_id in an authenticator may not contain
4829 binary zeroes or newlines because the value is written to log lines and
4830 to spool files. There was no check on this. Now the value is run through
4831 the string_printing() function so that such characters are converted to
4832 printable escape sequences.
4834 PH/41 $message_linecount is a new variable that contains the total number of
4835 lines in the message. Compare $body_linecount, which is the count for the
4838 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4839 bounce and delay warning messages, except in certain special cases, which
4842 (a) An SMTP error message from a remote host;
4843 (b) A message specified in a :fail: redirection;
4844 (c) A message specified in a "fail" command in a system filter;
4845 (d) A message specified in a FAIL return from the queryprogram router;
4846 (e) A message specified by the cannot_route_message router option.
4848 In these cases only, Exim does include the error details in bounce and
4849 warning messages. There are also a few cases where bland messages such
4850 as "unrouteable address" or "local delivery error" are given.
4852 PH/43 $value is now also set for the "else" part of a ${run expansion.
4854 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4855 being worked on, but at least Exim now implements the latest version to
4858 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4859 process to complete was treated as a delivery failure, a timeout while
4860 writing the message to the pipe was logged, but erroneously treated as a
4861 successful delivery. Such timeouts include transport filter timeouts. For
4862 consistency with the overall process timeout, these timeouts are now
4863 treated as errors, giving rise to delivery failures by default. However,
4864 there is now a new Boolean option for the pipe transport called
4865 timeout_defer, which, if set TRUE, converts the failures into defers for
4866 both kinds of timeout. A transport filter timeout is now identified in
4869 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4870 systems where "make" and "gmake" are different, calling "gmake" at top
4871 level broke things. I've arranged for the value of $(MAKE) to be passed
4872 from the Makefile to this script so that it can call the same version of
4876 A note about Exim versions 4.44 and 4.50
4877 ----------------------------------------
4879 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4880 changes of various kinds. As a consequence, a big documentation update was
4881 needed. This delayed the release for rather longer than seemed good, especially
4882 in the light of a couple of (minor) security issues. Therefore, the changes
4883 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4884 release. So 4.44 and 4.50 are in effect two different branches that both start
4887 I have left the 4.50 change log unchanged; it contains all the changes since
4888 4.43. The change log for 4.44 is below; many of its items are identical to
4889 those for 4.50. This seems to be the most sensible way to preserve the
4890 historical information.
4896 1. Minor wording change to the doc/README.SIEVE file.
4898 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4899 computation of the current number of files was incorrect.
4901 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4902 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4905 4. Give more explanation in the error message when the command for a transport
4906 filter fails to execute.
4908 5. There are several places where Exim runs a non-Exim command in a
4909 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4910 was being done only for the command run by the queryprogram router. It is
4911 now done for all such subprocesses. The other cases are: ${run, transport
4912 filters, and the commands run by the lmtp and pipe transports.
4914 6. Added CONFIGURE_GROUP build-time option.
4916 7. Some older OS have a limit of 256 on the maximum number of file
4917 descriptors. Exim was using setrlimit() to set 1000 as a large value
4918 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4919 systems. I've change it so that if it can't get 1000, it tries for 256.
4921 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4922 was an oversight, and furthermore, ever since the addition of extra
4923 controls (e.g. 4.43/32), the checks on when to allow different forms of
4924 "control" were broken. There should now be diagnostics for all cases when a
4925 control that does not make sense is encountered.
4927 9. Added the /retain_sender option to "control=submission".
4929 10. $recipients is now available in the predata ACL (oversight).
4931 11. Tidy the search cache before the fork to do a delivery from a message
4932 received from the command line. Otherwise the child will trigger a lookup
4933 failure and thereby defer the delivery if it tries to use (for example) a
4934 cached ldap connection that the parent has called unbind on.
4936 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4937 of $address_data from the recipient verification was clobbered by the
4938 sender verification.
4940 13. The value of address_data from a sender verification is now available in
4941 $sender_address_data in subsequent conditions in the ACL statement.
4943 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4945 15. Added a new option "connect=<time>" to callout options, to set a different
4948 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4949 was its contents. (It was OK if the option was not defined at all.)
4951 17. A "Completed" log line is now written for messages that are removed from
4952 the spool by the -Mrm option.
4954 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4955 information about exactly what failed.
4957 19. Added -dd to debug only the daemon process.
4959 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4960 handles timeouts, both on the server side and network timeouts. Renamed the
4961 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4963 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4964 It is now set to "smtps".
4966 22. $host_address is now set to the target address during the checking of
4967 ignore_target_hosts.
4969 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4970 being passed; this would have caused $sender_host_name to have been used if
4971 matching the list had actually called for a host name (not very likely,
4972 since this list is usually IP addresses). A host name is now passed as
4975 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4976 code that specifies a non-restarting handler (typically sigaction() in
4977 modern systems) in an attempt to fix a rare and obscure crash bug.
4979 25. Narrowed the window for a race in the daemon that could cause it to ignore
4980 SIGCHLD signals. This is not a major problem, because they are used only to
4981 wake it up if nothing else does.
4983 26. A malformed maildirsize file could cause Exim to calculate negative values
4984 for the mailbox size or file count. Odd effects could occur as a result.
4985 The maildirsize information is now recalculated if the size or filecount
4988 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4989 support for a long time. Removed HAVE_SYS_VFS_H.
4991 28. Installed the latest version of exipick from John Jetmore.
4993 29. In an address list, if the pattern was not a regular expression, an empty
4994 subject address (from a bounce message) matched only if the pattern was an
4995 empty string. Non-empty patterns were not even tested. This was the wrong
4996 because it is perfectly reasonable to use an empty address as part of a
4997 database query. An empty address is now tested by patterns that are
4998 lookups. However, all the other forms of pattern expect the subject to
4999 contain a local part and a domain, and therefore, for them, an empty
5000 address still always fails if the pattern is not itself empty.
5002 30. Exim went into a mad DNS loop when attempting to do a callout where the
5003 host was specified on an smtp transport, and looking it up yielded more
5004 than one IP address.
5006 31. Re-factored the code for checking spool and log partition space into a
5007 function that finds that data and another that does the check. The former
5008 is then used to implement four new variables: $spool_space, $log_space,
5009 $spool_inodes, and $log_inodes.
5011 32. The RFC2047 encoding function was originally intended for short strings
5012 such as real names; it was not keeping to the 75-character limit for
5013 encoded words that the RFC imposes. It now respects the limit, and
5014 generates multiple encoded words if necessary. To be on the safe side, I
5015 have increased the buffer size for the ${rfc2047: expansion operator from
5018 33. It is now permitted to omit both strings after an "if" condition; if the
5019 condition is true, the result is "true". As before, when the second string
5020 is omitted, a false condition yields an empty string. This makes it less
5021 cumbersome to write custom ACL and router conditions.
5023 34. Failure to deliver a bounce message always caused it to be frozen, even if
5024 there was an errors_to setting on the router. The errors_to setting is now
5027 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5028 canonical form (fully expanded) before being placed in
5029 $sender_host_address.
5031 36. The table in the code that translates DNS record types into text (T_A to
5032 "A" for instance) was missing entries for NS and CNAME. It is just possible
5033 that this could have caused confusion if both these types were looked up
5034 for the same domain, because the text type is used as part of Exim's
5035 per-process caching. But the chance of anyone hitting this buglet seems
5038 37. The dnsdb lookup has been extended in a number of ways.
5040 (1) There is a new type, "zns", which walks up the domain tree until it
5041 finds some nameserver records. It should be used with care.
5043 (2) There is a new type, "mxh", which is like "mx" except that it returns
5044 just the host names, not the priorities.
5046 (3) It is now possible to give a list of domains (or IP addresses) to be
5047 looked up. The behaviour when one of the lookups defers can be
5048 controlled by a keyword.
5050 (4) It is now possible to specify the separator character for use when
5051 multiple records are returned.
5053 38. The dnslists ACL condition has been extended: it is now possible to supply
5054 a list of IP addresses and/or domains to be looked up in a particular DNS
5057 39. Added log_selector=+queue_time_overall.
5059 40. When running the queue in the test harness, wait just a tad after forking a
5060 delivery process, to get repeatability of debugging output.
5062 41. Include certificate and key file names in error message when GnuTLS fails
5063 to set them up, because the GnuTLS error message doesn't include the name
5064 of the failing file when there is a problem reading it.
5066 42. Allow both -bf and -bF in the same test run.
5068 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5070 44. The "Exiscan patch" is now merged into the mainline Exim source.
5072 45. Sometimes the final signoff response after QUIT could fail to get
5073 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5074 < 0 before doing a fflush(). This bug looks as though it goes back to the
5075 introduction of TLS in release 3.20, but "sometimes" must have been rare
5076 because the tests only now provoked it.
5078 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5079 (this can affect the format of dates).
5081 47. exim_tidydb, when checking for the continued existence of a message for
5082 which it has found a message-specific retry record, was not finding
5083 messages that were in split spool directories. Consequently, it was
5084 deleting retry records that should have stayed in existence.
5086 48. Steve fixed some bugs in eximstats.
5088 49. The SPA authentication driver was not abandoning authentication and moving
5089 on to the next authenticator when an expansion was forced to fail,
5090 contradicting the general specification for all authenticators. Instead it
5091 was generating a temporary error. It now behaves as specified.
5093 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5094 (the order specifies the preference for clients). The order is now AES256,
5095 AES128, 3DES, ARCFOUR128.
5097 51. Small patch to Sieve code - explicitly set From: when generating an
5100 52. Exim crashed if a remote delivery caused a very long error message to be
5101 recorded - for instance if somebody sent an entire SpamAssassin report back
5102 as a large number of 550 error lines. This bug was coincidentally fixed by
5103 increasing the size of one of Exim's internal buffers (big_buffer) that
5104 happened as part of the Exiscan merge. However, to be on the safe side, I
5105 have made the code more robust (and fixed the comments that describe what
5108 53. Now that there can be additional text after "Completed" in log lines (if
5109 the queue_time_overall log selector is set), a one-byte patch to exigrep
5110 was needed to allow it to recognize "Completed" as not the last thing in
5113 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5114 patch that reportedly fixes this has been added. I am not expert enough to
5115 create a test for it. This is what the patch creator wrote:
5117 "I found a little strange behaviour of ldap code when working with
5118 Windows 2003 AD Domain, where users was placed in more than one
5119 Organization Units. When I tried to give exim partial DN, the exit code
5120 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5121 But simultaneously result of request was absolutely normal ldap result,
5122 so I produce this patch..."
5124 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5125 REFERENCE, so I have modified the code to exclude the patch when that macro
5128 55. Some experimental protocols are using DNS PTR records for new purposes. The
5129 keys for these records are domain names, not reversed IP addresses. The
5130 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5131 leaves it alone. Component reversal etc. now happens only for IP addresses.
5134 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5136 57. Double the size of the debug message buffer (to 2048) so that more of very
5137 long debug lines gets shown.
5139 58. The exicyclog utility now does better if the number of log files to keep
5140 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5142 59. Two changes related to the smtp_active_hostname option:
5144 (1) $smtp_active_hostname is now available as a variable.
5145 (2) The default for smtp_banner uses $smtp_active_hostname instead
5146 of $primary_hostname.
5148 60. The host_aton() function is supposed to be passed a string that is known
5149 to be a valid IP address. However, in the case of IPv6 addresses, it was
5150 not checking this. This is a hostage to fortune. Exim now panics and dies
5151 if the condition is not met. A case was found where this could be provoked
5152 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5153 components; fortuitously, this particular loophole had already been fixed
5154 by change 4.50/55 above.
5156 If there are any other similar loopholes, the new check in host_aton()
5157 itself should stop them being exploited. The report I received stated that
5158 data on the command line could provoke the exploit when Exim was running as
5159 exim, but did not say which command line option was involved. All I could
5160 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5161 running as the user.
5164 61. There was a buffer overflow vulnerability in the SPA authentication code
5165 (which came originally from the Samba project). I have added a test to the
5166 spa_base64_to_bits() function which I hope fixes it.
5169 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5170 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5172 63. The daemon start-up calls getloadavg() while still root for those OS that
5173 need the first call to be done as root, but it missed one case: when
5174 deliver_queue_load_max is set with deliver_drop_privilege. This is
5175 necessary for the benefit of the queue runner, because there is no re-exec
5176 when deliver_drop_privilege is set.
5178 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5179 This has been fixed.
5181 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5182 was in use, was not putting the data itself into the right store pool;
5183 consequently, it could be overwritten for a subsequent message in the same
5184 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5187 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5189 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5190 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5191 the function were treating the return as a boolean value, which happened to
5192 work because 0=false and not-0=true, but is not correct code.
5194 68. The host_aton() function was not handling scoped IPv6 addresses (those
5195 with, for example, "%eth0" on the end) correctly.
5197 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5198 negated items (that is, ~something) in unsigned ints. Some compilers
5199 apparently mutter when there is no cast.
5201 70. If an address verification called from an ACL failed, and did not produce a
5202 user-specific message (i.e. there was only a "system" message), nothing was
5203 put in $acl_verify_message. In this situation, it now puts the system
5206 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5207 session; change 4.31/43 added the unwanted input to the log line - except
5208 that it did not do this in the start of session case. It now does.
5210 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5211 This is wrong and can cause the other end to generate a synchronization
5212 error if it is another Exim or anything else that does the synchronization
5213 check. A QUIT command is no longer sent after a timeout.
5215 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5216 during host lookups.
5218 74. The defer_ok option of callout verification was not working if it was used
5219 when verifying addresses in header lines, that is, for this case:
5221 verify = header_sender/callout=defer_ok
5223 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5224 those file descriptors could be used for SMTP connections. If anything
5225 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5226 could be sent to the SMTP client, causing chaos. The daemon now opens
5227 stdin, stdout, and stderr to /dev/null when it puts itself into the
5230 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5231 log by default. The user can override this with suitable Perl magic.
5233 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5234 the log message when discard triggers, was not working for the DATA ACL or
5235 for the non-SMTP ACL.
5237 78. Error message wording change in sieve.c.
5239 79. If smtp_accept_max_per_host was set, the number of connections could be
5240 restricted to fewer than expected, because the daemon was trying to set up
5241 a new connection before checking whether the processes handling previous
5242 connections had finished. The check for completed processes is now done
5243 earlier. On busy systems, this bug wouldn't be noticed because something
5244 else would have woken the daemon, and it would have reaped the completed
5247 80. If a message was submitted locally by a user whose login name contained one
5248 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5249 It caused a spool format error. I have fixed the spool reading code. A
5250 related problem was that the "from" clause in the Received: line became
5251 illegal because of the space(s). It is now covered by ${quote_local_part.
5253 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5256 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5258 83. Patch from Sieve maintainer to fix unterminated string problem in
5259 "vacation" handling.
5261 84. Some minor changes to the Linux configuration files to help with other
5262 OS variants using glibc.
5264 85. One more patch for Sieve to update vacation handling to latest spec.
5267 ----------------------------------------------------
5268 See the note above about the 4.44 and 4.50 releases.
5269 ----------------------------------------------------
5275 1. Change 4.43/35 introduced a bug that caused file counts to be
5276 incorrectly computed when quota_filecount was set in an appendfile
5279 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5280 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5283 3. Give more explanation in the error message when the command for a transport
5284 filter fails to execute.
5286 4. There are several places where Exim runs a non-Exim command in a
5287 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5288 was being done only for the command run by the queryprogram router. It is
5289 now done for all such subprocesses. The other cases are: ${run, transport
5290 filters, and the commands run by the lmtp and pipe transports.
5292 5. Some older OS have a limit of 256 on the maximum number of file
5293 descriptors. Exim was using setrlimit() to set 1000 as a large value
5294 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5295 systems. I've change it so that if it can't get 1000, it tries for 256.
5297 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5298 was an oversight, and furthermore, ever since the addition of extra
5299 controls (e.g. 4.43/32), the checks on when to allow different forms of
5300 "control" were broken. There should now be diagnostics for all cases when a
5301 control that does not make sense is encountered.
5303 7. $recipients is now available in the predata ACL (oversight).
5305 8. Tidy the search cache before the fork to do a delivery from a message
5306 received from the command line. Otherwise the child will trigger a lookup
5307 failure and thereby defer the delivery if it tries to use (for example) a
5308 cached ldap connection that the parent has called unbind on.
5310 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5311 of $address_data from the recipient verification was clobbered by the
5312 sender verification.
5314 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5315 was its contents. (It was OK if the option was not defined at all.)
5317 11. A "Completed" log line is now written for messages that are removed from
5318 the spool by the -Mrm option.
5320 12. $host_address is now set to the target address during the checking of
5321 ignore_target_hosts.
5323 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5324 being passed; this would have caused $sender_host_name to have been used if
5325 matching the list had actually called for a host name (not very likely,
5326 since this list is usually IP addresses). A host name is now passed as
5329 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5330 code that specifies a non-restarting handler (typically sigaction() in
5331 modern systems) in an attempt to fix a rare and obscure crash bug.
5333 15. Narrowed the window for a race in the daemon that could cause it to ignore
5334 SIGCHLD signals. This is not a major problem, because they are used only to
5335 wake it up if nothing else does.
5337 16. A malformed maildirsize file could cause Exim to calculate negative values
5338 for the mailbox size or file count. Odd effects could occur as a result.
5339 The maildirsize information is now recalculated if the size or filecount
5342 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5343 support for a long time. Removed HAVE_SYS_VFS_H.
5345 18. Updated exipick to current release from John Jetmore.
5347 19. Allow an empty sender to be matched against a lookup in an address list.
5348 Previously the only cases considered were a regular expression, or an
5351 20. Exim went into a mad DNS lookup loop when doing a callout where the
5352 host was specified on the transport, if the DNS lookup yielded more than
5355 21. The RFC2047 encoding function was originally intended for short strings
5356 such as real names; it was not keeping to the 75-character limit for
5357 encoded words that the RFC imposes. It now respects the limit, and
5358 generates multiple encoded words if necessary. To be on the safe side, I
5359 have increased the buffer size for the ${rfc2047: expansion operator from
5362 22. Failure to deliver a bounce message always caused it to be frozen, even if
5363 there was an errors_to setting on the router. The errors_to setting is now
5366 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5367 canonical form (fully expanded) before being placed in
5368 $sender_host_address.
5370 24. Updated eximstats to version 1.33
5372 25. Include certificate and key file names in error message when GnuTLS fails
5373 to set them up, because the GnuTLS error message doesn't include the name
5374 of the failing file when there is a problem reading it.
5376 26. Expand error message when OpenSSL has problems setting up cert/key files.
5379 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5380 (this can affect the format of dates).
5382 28. exim_tidydb, when checking for the continued existence of a message for
5383 which it has found a message-specific retry record, was not finding
5384 messages that were in split spool directories. Consequently, it was
5385 deleting retry records that should have stayed in existence.
5387 29. eximstats updated to version 1.35
5388 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5389 1.35 - bugfix such that pie charts by volume are generated correctly
5391 30. The SPA authentication driver was not abandoning authentication and moving
5392 on to the next authenticator when an expansion was forced to fail,
5393 contradicting the general specification for all authenticators. Instead it
5394 was generating a temporary error. It now behaves as specified.
5396 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5397 (the order specifies the preference for clients). The order is now AES256,
5398 AES128, 3DES, ARCFOUR128.
5400 31. Small patch to Sieve code - explicitly set From: when generating an
5403 32. Exim crashed if a remote delivery caused a very long error message to be
5404 recorded - for instance if somebody sent an entire SpamAssassin report back
5405 as a large number of 550 error lines. This bug was coincidentally fixed by
5406 increasing the size of one of Exim's internal buffers (big_buffer) that
5407 happened as part of the Exiscan merge. However, to be on the safe side, I
5408 have made the code more robust (and fixed the comments that describe what
5411 33. Some experimental protocols are using DNS PTR records for new purposes. The
5412 keys for these records are domain names, not reversed IP addresses. The
5413 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5414 leaves it alone. Component reversal etc. now happens only for IP addresses.
5417 34. The host_aton() function is supposed to be passed a string that is known
5418 to be a valid IP address. However, in the case of IPv6 addresses, it was
5419 not checking this. This is a hostage to fortune. Exim now panics and dies
5420 if the condition is not met. A case was found where this could be provoked
5421 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5422 components; fortuitously, this particular loophole had already been fixed
5423 by change 4.50/55 or 4.44/33 above.
5425 If there are any other similar loopholes, the new check in host_aton()
5426 itself should stop them being exploited. The report I received stated that
5427 data on the command line could provoke the exploit when Exim was running as
5428 exim, but did not say which command line option was involved. All I could
5429 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5430 running as the user.
5433 35. There was a buffer overflow vulnerability in the SPA authentication code
5434 (which came originally from the Samba project). I have added a test to the
5435 spa_base64_to_bits() function which I hope fixes it.
5438 36. The daemon start-up calls getloadavg() while still root for those OS that
5439 need the first call to be done as root, but it missed one case: when
5440 deliver_queue_load_max is set with deliver_drop_privilege. This is
5441 necessary for the benefit of the queue runner, because there is no re-exec
5442 when deliver_drop_privilege is set.
5444 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5445 was in use, was not putting the data itself into the right store pool;
5446 consequently, it could be overwritten for a subsequent message in the same
5447 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5450 38. Sometimes the final signoff response after QUIT could fail to get
5451 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5452 < 0 before doing a fflush(). This bug looks as though it goes back to the
5453 introduction of TLS in release 3.20, but "sometimes" must have been rare
5454 because the tests only now provoked it.
5460 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5461 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5462 Now it returns an integer. A number of calls were still expecting a T/F
5463 return. Fortuitously, in all cases, the tests worked in OK situations,
5464 which is the norm. However, things would have gone wrong on any write
5465 failures on the smtp file descriptor. This function is used when sending
5466 messages over SMTP and also when doing verify callouts.
5468 2. When Exim is called to do synchronous delivery of a locally submitted
5469 message (the -odf or -odi options), it no longer closes stderr before doing
5472 3. Implemented the mua_wrapper option.
5474 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5476 5. Implemented the functions header_remove(), header_testname(),
5477 header_add_at_position(), and receive_remove_recipient(), and exported them
5480 6. If an ACL "warn" statement specified the addition of headers, Exim already
5481 inserted X-ACL-Warn: at the start if there was no header name. However, it
5482 was not making this test for the second and subsequent header lines if
5483 there were newlines in the string. This meant that an invalid header could
5484 be inserted if Exim was badly configured.
5486 7. Allow an ACL "warn" statement to add header lines at the start or after all
5487 the Received: headers, as well as at the end.
5489 8. Added the rcpt_4xx retry error code.
5491 9. Added postmaster_mailfrom=xxx to callout verification option.
5493 10. Added mailfrom=xxxx to the callout verification option, for verify=
5496 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5497 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5499 12. Inserted some casts to stop certain compilers warning when using pointer
5500 differences as field lengths or precisions in printf-type calls (mostly
5501 affecting debugging statements).
5503 13. Added optional readline() support for -be (dynamically loaded).
5505 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5506 same clock tick as a message's arrival, so that its received time was the
5507 same as the "first fail" time on the retry record, and that message
5508 remained on the queue past the ultimate address timeout, every queue runner
5509 would try a delivery (because it was past the ultimate address timeout) but
5510 after another failure, the ultimate address timeout, which should have then
5511 bounced the address, did not kick in. This was a "< instead of <=" error;
5512 in most cases the first failure would have been in the next clock tick
5513 after the received time, and all would be well.
5515 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5516 being recognized when the domain list was tested by the match_domain
5517 condition in an expansion string.
5519 16. Added the ${str2b64: operator.
5521 17. Exim was always calling setrlimit() to set a large limit for the number of
5522 processes, without checking whether the existing limit was already
5523 adequate. (It did check for the limit on file descriptors.) Furthermore,
5524 errors from getrlimit() and setrlimit() were being ignored. Now they are
5525 logged to the main and panic logs, but Exim does carry on, to try to do its
5526 job under whatever limits there are.
5528 18. Imported PCRE 5.0.
5530 19. Trivial typo in log message " temporarily refused connection" (the leading
5533 20. If the log selector return_path_on_delivery was set and an address was
5534 redirected to /dev/null, the delivery process crashed because it assumed
5535 that a return path would always be set for a "successful" delivery. In this
5536 case, the whole delivery is bypassed as an optimization, and therefore no
5539 21. Internal re-arrangement: the function for sending a challenge and reading
5540 a response while authentication was assuming a zero-terminated challenge
5541 string. It's now changed to take a pointer and a length, to allow for
5542 binary data in such strings.
5544 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5546 23. Exim was not respecting finduser_retries when seeking the login of the
5547 uid under which it was called; it was always trying 10 times. (The default
5548 setting of finduser_retries is zero.) Also, it was sleeping after the final
5549 failure, which is pointless.
5551 24. Implemented tls_on_connect_ports.
5553 25. Implemented acl_smtp_predata.
5555 26. If the domain in control=submission is set empty, Exim assumes that the
5556 authenticated id is a complete email address when it generates From: or
5557 Sender: header lines.
5559 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5560 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5561 chown and chgrp in /bin and hostname in /usr/bin.
5563 28. Exim was keeping the "process log" file open after each use, just as it
5564 does for the main log. This opens the possibility of it remaining open for
5565 long periods when the USR1 signal hits a daemon. Occasional processlog
5566 errors were reported, that could have been caused by this. Anyway, it seems
5567 much more sensible not to leave this file open at all, so that is what now
5570 29. The long-running daemon process does not normally write to the log once it
5571 has entered its main loop, and it closes the log before doing so. This is
5572 so that log files can straightforwardly be renamed and moved. However,
5573 there are a couple of unusual error situations where the daemon does write
5574 log entries, and I had neglected to close the log afterwards.
5576 30. The text of an SMTP error response that was received during a remote
5577 delivery was being truncated at 512 bytes. This is too short for some of
5578 the long messages that one sometimes sees. I've increased the limit to
5581 31. It is now possible to make retry rules that apply only when a message has a
5582 specific sender, in particular, an empty sender.
5584 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5585 it possible to be selective about when SMTP synchronization is enforced.
5587 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5589 32. Implemented hosts_connection_nolog.
5591 33. Added an ACL for QUIT.
5593 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5596 35. Added mailbox_size and mailbox_filecount to appendfile.
5598 36. Added control = no_multiline_responses to ACLs.
5600 37. There was a bug in the logic of the code that waits for the clock to tick
5601 in the case where the clock went backwards by a substantial amount such
5602 that the microsecond fraction of "now" was more than the microsecond
5603 fraction of "then" (but the whole seconds number was less).
5605 38. Added support for the libradius Radius client library this is found on
5606 FreeBSD (previously only the radiusclient library was supported).
5612 1. When certain lookups returned multiple values in the form name=value, the
5613 quoting of the values was not always being done properly. Specifically:
5614 (a) If the value started with a double quote, but contained no whitespace,
5616 (b) If the value contained whitespace other than a space character (i.e.
5617 tabs or newlines or carriage returns) it was not quoted.
5618 This fix has been applied to the mysql and pgsql lookups by writing a
5619 separate quoting function and calling it from the lookup code. The fix
5620 should probably also be applied to nisplus, ibase and oracle lookups, but
5621 since I cannot test any of those, I have not disturbed their existing code.
5623 2. A hit in the callout cache for a specific address caused a log line with no
5624 reason for rejecting RCPT. Now it says "Previous (cached) callout
5625 verification failure".
5627 3. There was an off-by-one bug in the queryprogram router. An over-long
5628 return line was truncated at 256 instead of 255 characters, thereby
5629 overflowing its buffer with the terminating zero. As well as fixing this, I
5630 have increased the buffer size to 1024 (and made a note to document this).
5632 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5633 when Exim is waiting for an SMTP response from a remote server, Exim
5634 restarts its select() call on the socket, thereby resetting its timeout.
5635 This is not a problem when such interrupts are rare. Somebody set up a cron
5636 job to run exiwhat every 2 minutes, which is less than the normal select()
5637 timeout (5 or 10 minutes). This meant that the select() timeout never
5638 kicked in because it was always reset. I have fixed this by comparing the
5639 time when an interrupt arrives with the time at the start of the first call
5640 to select(). If more time than the timeout has elapsed, the interrupt is
5641 treated as a timeout.
5643 5. Some internal re-factoring in preparation for the addition of Sieve
5644 extensions (by MH). In particular, the "personal" test is moved to a
5645 separate function, and given an option for scanning Cc: and Bcc: (which is
5646 not set for Exim filters).
5648 6. When Exim created an email address using the login of the caller as the
5649 local part (e.g. when creating a From: or Sender: header line), it was not
5650 quoting the local part when it contained special characters such as @.
5652 7. Installed new OpenBSD configuration files.
5654 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5655 try to make them clearer.
5657 9. Callout options, other than the timeout value, were being ignored when
5658 verifying sender addresses in header lines. For example, when using
5660 verify = header_sender/callout=no_cache
5662 the cache was (incorrectly) being used.
5664 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5665 only the Cygwin environment.
5667 11. When return_path_on_delivery was set as a log selector, if different remote
5668 addresses in the same message used different return paths and parallel
5669 remote delivery occurred, the wrong values would sometimes be logged.
5670 (Whenever a remote delivery process finished, the return path value from
5671 the most recently started remote delivery process was logged.)
5673 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5674 lines when AUTH and/or TLS are in use. This is the "received protocol"
5675 field. Exim used to use "asmtp" for authenticated SMTP, without any
5676 indication (in the protocol name) for TLS use. Now it follows the RFC and
5677 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5678 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5679 names appear in log lines as well as in Received: header lines.
5681 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5682 extensions, and comparison tests, and to fix some bugs.
5684 14. Changes to the "personal" filter test:
5686 (1) The test was buggy in that it was just doing the equivalent of
5687 "contains" tests on header lines. For example, if a user's address was
5688 anne@some.where, the "personal" test would incorrectly be true for
5690 To: susanne@some.where
5692 This test is now done by extracting each address from the header in turn,
5693 and checking the entire address. Other tests that are part of "personal"
5694 are now done using regular expressions (for example, to check local parts
5695 of addresses in From: header lines).
5697 (2) The list of non-personal local parts in From: addresses has been
5698 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5699 taken from the Sieve specification recommendations.
5701 (3) If the message contains any header line starting with "List-" it is
5702 treated as non-personal.
5704 (4) The test for "circular" in the Subject: header line has been removed
5705 because it now seems ill-conceived.
5707 15. Minor typos in src/EDITME comments corrected.
5709 16. Installed latest exipick from John Jetmore.
5711 17. If headers_add on a router specified a text string that was too long for
5712 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5713 of string_sprintf() is now avoided.
5715 18. $message_body_size was not set (it was always zero) when running the DATA
5716 ACL and the local_scan() function.
5718 19. For the "mail" command in an Exim filter, no default was being set for
5719 the once_repeat time, causing a random time value to be used if "once" was
5720 specified. (If the value happened to be <= 0, no repeat happened.) The
5721 default is now 0s, meaning "never repeat". The "vacation" command was OK
5722 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5723 (I found it when inspecting the code).
5725 20. There is now an overall timeout for performing a callout verification. It
5726 defaults to 4 times the callout timeout, which applies to individual SMTP
5727 commands during the callout. The overall timeout applies when there is more
5728 than one host that can be tried. The timeout is checked before trying the
5729 next host. This prevents very long delays if there are a large number of
5730 hosts and all are timing out (e.g. when the network connections are timing
5731 out). The value of the overall timeout can be changed by specifying an
5732 additional sub-option for "callout", called "maxwait". For example:
5734 verify = sender/callout=5s,maxwait=20s
5736 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5737 to the end before writing, but this should make it even safer).
5739 22. Exim was forgetting that it had advertised PIPELINING for the second and
5740 subsequent messages on an SMTP connection. It was also not resetting its
5741 memory on STARTTLS and an internal HELO.
5743 23. When Exim logs an SMTP synchronization error within a session, it now
5744 records whether PIPELINING has been advertised or not.
5746 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5747 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5748 rather than long int.
5750 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5752 26. Added the never_mail option to autoreply.
5758 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5759 crash if the getsockname() call failed; this can happen if a connection is
5760 closed very soon after it is established. The problem was simply in the
5761 order in which certain operations were done, causing Exim to try to write
5762 to the SMTP stream before it had set up the file descriptor. The bug has
5763 been fixed by making things happen in the correct order.
5769 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5770 before the connection was closed, thus losing the rejection response.
5772 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5773 some early Solaris releases, but causes trouble in current releases where
5774 socklen_t is defined.
5776 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5779 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5782 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5783 that releasing the top of it at the end releases what was used for sub-
5784 expansions (unless the block got too big). However, discard this block if
5785 the first thing is a variable or header, so that we can use its block when
5786 it is dynamic (useful for very large $message_headers, for example).
5788 6. Lookups now cache *every* query, not just the most recent. A new, separate
5789 store pool is used for this. It can be recovered when all lookup caches are
5790 flushed. Lookups now release memory at the end of their result strings.
5791 This has involved some general refactoring of the lookup sources.
5793 7. Some code has been added to the store_xxx() functions to reduce the amount
5794 of flapping under certain conditions.
5796 8. log_incoming_interface used to affect only the <= reception log lines. Now
5797 it causes the local interface and port to be added to several more SMTP log
5798 lines, for example "SMTP connection from", and rejection lines.
5800 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5802 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5804 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5805 could be overwritten at the end of the current message (or the start of a
5806 new message if it was set in a HELO ACL). The value is now preserved for
5807 the duration of the SMTP connection.
5809 12. If a transport had a headers_rewrite setting, and a matching header line
5810 contained an unqualified address, that address was qualified, even if it
5811 did not match any rewriting rules. The underlying bug was that the values
5812 of the flags that permit the existence of unqualified sender and recipient
5813 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5814 non-local messages, and by -bnq for local messages) were not being
5815 preserved with the message after it was received.
5817 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5818 "next input=" as part of the text comprising the host identity instead of
5819 the correct text. The code was using the same buffer for two different
5820 strings. However, depending on which order the printing function evaluated
5821 its arguments, the bug did not always show up. Under Linux, for example, my
5822 test suite worked just fine.
5824 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5825 doesn't work with some older versions of Perl. It has been changed to "my",
5826 which in any case is probably the better facility to use.
5828 15. A really picky compiler found some instances of statements for creating
5829 error messages that either had too many or two few arguments for the format
5832 16. The size of the buffer for calls to the DNS resolver has been increased
5833 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5834 for addresses that have a lot of PTR records. This alleviates a problem; it
5835 does not fully solve it.
5837 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5838 buffer now truncates the list and logs the incident, which is the same
5839 action as happens when Exim is looking up a host name and its aliases.
5840 Previously in this situation something unpredictable would happen;
5841 sometimes it was "internal error: store_reset failed".
5843 18. If a server dropped the connection unexpectedly when an Exim client was
5844 using GnuTLS and trying to read a response, the client delivery process
5845 crashed while trying to generate an error log message.
5847 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5848 string, for example:
5850 warn message = H1: something\nH2: something
5852 the text was added as a single header line from Exim's point of view
5853 though it ended up OK in the delivered message. However, searching for the
5854 second and subsequent header lines using $h_h2: did not work. This has been
5855 fixed. Similarly, if a system filter added multiple headers in this way,
5856 the routers could not see them.
5858 20. Expanded the error message when iplsearch is called with an invalid key to
5859 suggest using net-iplsearch in a host list.
5861 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5862 ACLs are no longer actually imposed (and a message to that effect is
5865 22. If a "gecos" field in a passwd entry contained escaped characters, in
5866 particular, if it contained a \" sequence, Exim got it wrong when building
5867 a From: or a Sender: header from that name. A second bug also caused
5868 incorrect handling when an unquoted " was present following a character
5869 that needed quoting.
5871 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5872 was not being matched caselessly.
5874 24. Arranged for all hyphens in the exim.8 source to be escaped with
5877 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5878 properly. Recipient callout cache records were still being keyed to include
5879 the sender, even when use_sender was set false. This led to far more
5880 callouts that were necessary. The sender is no longer included in the key
5881 when use_sender is false.
5883 26. Added "control = submission" modifier to ACLs.
5885 27. Added the ${base62d: operator to decode base 62 numbers.
5887 28. dnsdb lookups can now access SRV records.
5889 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5890 the configuration file.
5892 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5893 (-v) mode. This makes the output for a verbose queue run more intelligible.
5895 31. Added a use_postmaster feature to recipient callouts.
5897 32. Added the $body_zerocount variable, containing the number of binary zero
5898 bytes in the message body.
5900 33. The time of last modification of the "new" subdirectory is now used as the
5901 "mailbox time last read" when there is a quota error for a maildir
5904 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5906 35. Added +ignore_unknown as a special item in host lists.
5908 36. Code for decoding IPv6 addresses in host lists is now included, even if
5909 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5910 address was recognized as an IP address, but was then not correctly decoded
5911 into binary, causing unexpected and incorrect effects when compared with
5918 1. Very minor rewording of debugging text in manualroute to say "list of
5919 hosts" instead of "hostlist".
5921 2. If verify=header_syntax was set, and a header line with an unqualified
5922 address (no domain) and a large number of spaces between the end of the
5923 name and the colon was received, the reception process suffered a buffer
5924 overflow, and (when I tested it) crashed. This was caused by some obsolete
5925 code that should have been removed. The fix is to remove it!
5927 3. When running in the test harness, delay a bit after writing a bounce
5928 message to get a bit more predictability in the log output.
5930 4. Added a call to search_tidyup() just before forking a reception process. In
5931 theory, someone could use a lookup in the expansion of smtp_accept_max_
5932 per_host which, without the tidyup, could leave open a database connection.
5934 5. Added the variables $recipient_data and $sender_data which get set from a
5935 lookup success in an ACL "recipients" or "senders" condition, or a router
5936 "senders" option, similar to $domain_data and $local_part_data.
5938 6. Moved the writing of debug_print from before to after the "senders" test
5941 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5942 problems for message scanning, either using a data ACL, or using
5943 local_scan() because the Received: header was not generated till after they
5944 were called (in order to set the time as the time of reception completion).
5945 I have revised the way this works. The header is now generated after the
5946 body is received, but before the ACL or local_scan() are called. After they
5947 are run, the timestamp in the header is updated.
5953 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5954 before starting a queue runner without re-exec. This happened only when
5955 deliver_drop_privilege was set or when the Exim user was set to root. The
5956 effect of the bug was that timeouts during subsequent deliveries caused
5957 crashes instead of being properly handled. The handler is now left at its
5958 default (and expected) setting.
5960 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5961 message, again when deliver_drop_privilege is set or Exim is run as root.
5962 The bug described in (1) was not present in this case, but the tidying up
5963 of the other signals was missing. I have made the two cases consistent.
5965 3. The ignore_target_hosts setting on a manualroute router was being ignored
5966 for hosts that were looked up using the /MX notation.
5968 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5971 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5972 operated on the sender address. After changing the $sender_address to <>
5973 for the sender address verify, Exim was re-instated it as the original
5974 (before rewriting) address, but remembering that it had rewritten it, so it
5975 wasn't rewriting it again. This bug also had the effect of breaking the
5976 sender address verification caching when the sender address was rewritten.
5978 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5979 This has been changed so that if the ip literal address matches
5980 ignore_target_hosts, the router declines.
5982 7. Added expansion conditions match_domain, match_address, and match_local_
5983 part (NOT match_host).
5985 8. The placeholder for the Received: header didn't have a length field set.
5987 9. Added code to Exim itself and to exim_lock to test for a specific race
5988 condition that could lead to file corruption when using MBX delivery. The
5989 issue is with the lockfile that is created in /tmp. If this file is removed
5990 after a process has opened it but before that process has acquired a lock,
5991 there is the potential for a second process to recreate the file and also
5992 acquire a lock. This could lead to two Exim processes writing to the file
5993 at the same time. The added code performs the same test as UW imapd; it
5994 checks after acquiring the lock that its file descriptor still refers to
5995 the same named file.
5997 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5998 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6001 11. Added the smtp_active_hostname option. If used, this will typically be made
6002 to depend on the incoming interface address. Because $interface_address is
6003 not set up until the daemon has forked a reception process, error responses
6004 that can happen earlier (such as "too many connections") no longer contain
6007 12. If an expansion in a condition on a "warn" statement fails because a lookup
6008 defers, the "warn" statement is abandoned, and the next ACL statement is
6009 processed. Previously this caused the whole ACL to be aborted.
6011 13. Added the iplsearch lookup type.
6013 14. Added ident_timeout as a log selector.
6015 15. Added tls_certificate_verified as a log selector.
6017 16. Added a global option tls_require_ciphers (compare the smtp transport
6018 option of the same name). This controls incoming TLS connections.
6020 17. I finally figured out how to make tls_require_ciphers do a similar thing
6021 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6022 before starting the TLS session.
6024 18. Tabs are now shown as \t in -bP output.
6026 19. If the log selector return_path_on_delivery was set, Exim crashed when
6027 bouncing a message because it had too many Received: header lines.
6029 20. If two routers both had headers_remove settings, and the first one included
6030 a superfluous trailing colon, the final name in the first list and the
6031 first name in the second list were incorrectly joined into one item (with a
6032 colon in the middle).
6038 1. Added -C and -D options to the exinext utility, mainly to make it easier
6039 to include in the automated testing, but these could be helpful when
6040 multiple configurations are in use.
6042 2. The exinext utility was not formatting the output nicely when there was
6043 an alternate port involved in the retry record key, nor when there was a
6044 message id as well (for retries that were specific to a specific message
6045 and a specific host). It was also confused by IPv6 addresses, because of
6046 the additional colons they contain. I have fixed the IPv4 problem, and
6047 patched it up to do a reasonable job for IPv6.
6049 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6050 delivery, the log line now contains "pipelined" if PIPELINING was used.
6052 4. An SMTP transport process used to panic and die if the bind() call to set
6053 an explicit outgoing interface failed. This has been changed; it is now
6054 treated in the same way as a connect() failure.
6056 5. A reference to $sender_host_name in the part of a conditional expansion
6057 that was being skipped was still causing a DNS lookup. This no longer
6060 6. The def: expansion condition was not recognizing references to header lines
6061 that used bh_ and bheader_.
6063 7. Added the _cache feature to named lists.
6065 8. The code for checking quota_filecount in the appendfile transport was
6066 allowing one more file than it should have been.
6068 9. For compatibility with Sendmail, the command line option
6076 and sets the incoming protocol and host name (for trusted callers). The
6077 host name and its colon can be omitted when only the protocol is to be set.
6078 Note the Exim already has two private options, -pd and -ps, that refer to
6079 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6080 "s", but I don't think that's a major issue.
6082 10. A number of refactoring changes to the code, none of which should affect
6085 (a) The number of logging options was getting close to filling up the
6086 32-bit word that was used as a bit map. I have split them into two classes:
6087 those that are passed in the argument to log_write(), and those that are
6088 only ever tested independently outside of that function. These are now in
6089 separate 32-bit words, so there is plenty of room for expansion again.
6090 There is no change in the user interface or the logging behaviour.
6092 (b) When building, for example, log lines, the code previously used a
6093 macro that called string_cat() twice, in order to add two strings. This is
6094 not really sufficiently general. Furthermore, there was one instance where
6095 it was actually wrong because one of the argument was used twice, and in
6096 one call a function was used. (As it happened, calling the function twice
6097 did not affect the overall behaviour.) The macro has been replaced by a
6098 function that can join an arbitrary number of extra strings onto a growing
6101 (c) The code for expansion conditions now uses a table and a binary chop
6102 instead of a serial search (which was left over from when there were very
6103 few conditions). Also, it now recognizes conditions like "pam" even when
6104 the relevant support is not compiled in: a suitably worded error message is
6105 given if an attempt is made to use such a condition.
6107 11. Added ${time_interval:xxxxx}.
6109 12. A bug was causing one of the ddress fields not to be passed back correctly
6110 from remote delivery subprocesses. The field in question was not being
6111 subsequently used, so this caused to problems in practice.
6113 13. Added new log selectors queue_time and deliver_time.
6115 14. Might have fixed a bug in maildirsizefile handling that threw up
6116 "unexpected character" debug warnings, and recalculated the data
6117 unnecessarily. In any case, I expanded the warning message to give more
6120 15. Added the message "Restricted characters in address" to the statements in
6121 the default ACL that block characters like @ and % in local parts.
6123 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6124 Three changes have been made:
6126 (a) There was a serious bug; a negative response to MAIL caused the whole
6127 recipient domain to be cached as invalid, thereby blocking all messages
6128 to all local parts at the same domain, from all senders. This bug has
6129 been fixed. The domain is no longer cached after a negative response to
6130 MAIL if the sender used is not empty.
6132 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6135 (c) A new callout option, "use_sender" has been added for people who want
6136 the modified behaviour.
6142 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6145 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6146 indeed breaks things for older releases.
6148 3. Added additional logging to the case where there is a problem reading data
6149 from a filter that is running in a subprocess using a pipe, in order to
6150 try to track down a specific problem.
6152 4. Testing facility fudge: when running in the test harness and attempting
6153 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6154 getting "No route to host". Convert this to a timeout.
6156 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6159 6. Some OS don't have socklen_t but use size_t instead. This affects the
6160 fifth argument of getsockopt() amongst other things. This is now
6161 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6162 can be set for individual OS. I have set it for SunOS5, OSF1, and
6163 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6164 some earlier ones do not.
6166 7. Change 4.30/15 was not doing the test caselessly.
6168 8. The standard form for an IPv6 address literal was being rejected by address
6169 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6170 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6171 this, as well as the form without the "IPv6" on the front (but only when
6172 address literals are enabled, of course).
6174 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6176 10. Exim crashed if a message with an empty sender address specified by -f
6177 encountered a router with an errors_to setting. This could be provoked only
6178 by a command such as
6182 where an empty string was supplied; "<>" did not hit this bug.
6184 11. Installed PCRE release 4.5.
6186 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6187 remained set. It is now erased.
6189 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6190 times from message ids (which are base 36 rather than the normal 62).
6192 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6193 were being counted as actual protocol errors, and logged if the log
6194 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6195 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6196 and DATA following a set of rejected RCPTs do not count as protocol errors.
6197 In other words, Exim assumes they were pipelined, though this may not
6198 actually be the case. Of course, in all cases the client gets an
6199 appropriate error code.
6201 15. If a lookup fails in an ACL condition, a message about the failure may
6202 be available; it is used if testing the ACL cannot continue, because most
6203 such messages specify what the cause of the deferral is. However, some
6204 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6205 that caused an old message to be retained and used if a later statement
6206 caused a defer, replacing the real cause of the deferral.
6208 16. If an IP address had so many PTR records that the DNS lookup buffer
6209 was not large enough to hold them, Exim could crash while trying to process
6210 the truncated data. It now detects and logs this case.
6212 17. Further to 4.21/58, another change has been made: if (and only if) the
6213 first line of a message (the first header line) ends with CRLF, a bare LF
6214 in a subsequent header line has a space inserted after it, so as not to
6215 terminate the header.
6217 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6218 unnecessarily, used atoi() instead of strtol(), and didn't check the
6219 termination when getting file sizes from file names by regex.
6221 19. Completely re-implemented the support for maildirsize files, in the light
6222 of a number of problems with the previous contributed implementation
6223 (4.30/29). In particular:
6225 . If the quota is zero, the maildirsize file is maintained, but no quota is
6228 . If the maildir directory does not exist, it is created before any attempt
6229 to write a maildirsize file.
6231 . The quota value in the file is just a cache; if the quota is changed in
6232 the transport, the new value overrides.
6234 . A regular expression is available for excluding directories from the
6237 20. The autoreply transport checks the characters in options that define the
6238 message's headers; it allows continued headers, but it was checking with
6239 isspace() after an embedded newline instead of explicitly looking for a
6242 21. If all the "regular" hosts to which an address was routed had passed their
6243 expiry times, and had not reached their retry times, the address was
6244 bounced, even if fallback hosts were defined. Now Exim should go on to try
6247 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6248 equivalent code in the SMTP transport. Some hosts send humungous responses
6249 to HELO/EHLO, more than 1024 it seems.
6251 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6252 gives compiler warnings in some environments. I've now done it "properly",
6255 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6256 (because of problems with the built-in one) was declared to return uschar *
6257 instead of char *, causing compiler failure.
6259 25. Fixed a file descriptor leak when processing alias/forward files.
6261 26. Fixed a minor format string issue in dbfn.c.
6263 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6265 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6266 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6269 29. When a sender address is verified, it is cached, to save repeating the test
6270 when there is more than one recipient in a message. However, when the
6271 verification involves a callout, it is possible for different callout
6272 options to be set for different recipients. It is too complicated to keep
6273 track of this in the cache, so now Exim always runs a verification when a
6274 callout is required, relying on the callout cache for the optimization.
6275 The overhead is duplication of the address routing, but this should not be
6278 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6279 to be verified with callout=postmaster, and the main callout worked but the
6280 postmaster check failed, the verification correctly failed. However, if a
6281 subsequent RCPT command asked for sender verification *without* the
6282 postmaster check, incorrect caching caused this verification also to fail,
6285 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6286 it was not caching the DNS options (qualify_single, search_parents) that
6287 were used when the lookup failed. A subsequent lookup with different
6288 options therefore always gave the same answer, though there were cases
6289 where it should not have. (Example: a "domains = !$mx_any" option on a
6290 dnslookup router: the "domains" option is always processed without any
6291 widening, but the router might have qualify_single set.) Now Exim uses the
6292 cached value only when the same options are set.
6294 32. Added John Jetmore's "exipick" utility to the distribution.
6296 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6297 than a timeout (e.g. a certificate is required, and is not provided), an
6298 Exim server now closes the connection immediately. Previously it waited for
6299 the client to close - but if the client is SSL, it seems that they each
6300 wait for each other, leading to a delay before one of them times out.
6302 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6303 maintained 0.8.x compatibility because I don't think many are using it, and
6304 it is clearly obsolete.
6306 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6309 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6310 client certificate was expired. A simple patch fixes this, though I don't
6311 understand the full logic of why the verify callback is called multiple
6314 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6315 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6316 which causes problems with some clients (such as the Certicom SSL Plus
6317 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6318 disables the coutermeasure allowing Eudora to connect."
6320 38. Exim was not checking that a write() to a log file succeeded. This could
6321 lead to Bad Things if a log got too big, in particular if it hit a file
6322 size limit. Exim now panics and dies if it cannot write to a log file, just
6323 as it does if it cannot open a log file.
6325 39. Modified OS/Makefile-Linux so that it now contains
6327 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6329 The two -D definitions ensure that Exim is compiled with large file
6330 support, which makes it possible to handle log files that are bigger than
6333 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6334 instance) a domain was checked against a named list that involved a lookup,
6335 causing $domain_data to be set, then another domain was checked against the
6336 same list, then the first domain was re-checked, the value of $domain_data
6337 after the final check could be wrong. In particular, if the second check
6338 failed, it could be set empty. This bug probably also applied to
6341 41. The strip_trailing_dot option was not being applied to the address given
6342 with the -f command-line option.
6344 42. The code for reading a message's header from the spool was incrementing
6345 $received_count, but never initializing it. This meant that the value was
6346 incorrect (doubled) while delivering a message in the same process in which
6347 it was received. In the most common configuration of Exim, this never
6348 happens - a fresh exec is done - but it can happen when
6349 deliver_drop_privilege is set.
6351 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6352 it now includes up to 150 characters of the unexpected data in the log
6355 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6356 and building data strings. The size of both of these buffers was 10 000
6357 bytes - far larger than anybody would *ever* want, thought I. Needless to
6358 say, somebody hit the limit. I have increased the maximum line length to
6359 20 000 and the maximum data length of concatenated lines to 100 000. I have
6360 also fixed two bugs, because there was no checking on these buffers. Tsk,
6361 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6362 buffer is too small.
6364 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6365 lsearch lookups. Now it does.
6367 46. When parsing a route_list item in a manualroute router, a fixed-length
6368 buffer was used for the list of hosts. I made this 1024 bytes long,
6369 thinking that nobody would ever have a list of hosts that long. Wrong.
6370 Somebody had a whole pile of complicated expansion conditions, and the
6371 string was silently truncated, leading to an expansion error. It turns out
6372 that it is easier to change to an unlimited length (owing to other changes
6373 that have happened since this code was originally written) than to build
6374 structure for giving a limitation error. The length of the item that
6375 expands into the list of hosts is now unlimited.
6377 47. The lsearch lookup could not handle data where the length of text line was
6378 more than 4095 characters. Such lines were truncated, leading to shortened
6379 data being returned. It should now handle lines of any length.
6381 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6382 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6385 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6386 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6387 no longer happens in comment lines. A list of replacements is now placed
6388 at the head of all of the source files, except those whose only change is
6389 to replace PERL_COMMAND in the very first #! line.
6391 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6392 messages on the queue, with a bottom-up merge sort, using code contributed
6393 by Michael Haardt. This should make operations like -bp somewhat faster on
6394 large queues. It won't affect queue runners, except when queue_run_in_order
6397 51. Installed eximstats 1.31 in the distribution.
6399 52. Added support for SRV lookups to the dnslookup router.
6401 53. If an ACL referred to $message_body or $message_body_end, the value was not
6402 reset for any messages that followed in the same SMTP session.
6404 54. The store-handling optimization for building very long strings was not
6405 differentiating between the different store pools. I don't think this
6406 actually made any difference in practice, but I've tidied it.
6408 55. While running the routers to verify a sender address, $sender_address
6409 was still set to the sender address. This is wrong, because when routing to
6410 send a bounce to the sender, it would be empty. Therefore, I have changed
6411 it so that, while verifying a sender address, $sender_address is set to <>.
6412 (There is no change to what happens when verifying a recipient address.)
6414 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6415 target A or AAAA records (if not already returned) without resetting the
6416 qualify_single or search_parents options of the DNS resolver. These are
6417 inappropriate in this case because the targets of MX and SRV records must
6418 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6419 target that, when qualified, matched something in the local domain. These
6420 two options are now turned off when doing these lookups.
6422 57. It seems that at least some releases of Reiserfs (which does not have the
6423 concept of a fixed number of inodes) returns zero and not -1 for the
6424 number of available inodes. This interacted badly with check_spool_inodes,
6425 which assumed that -1 was the "no such thing" setting. What I have done is
6426 to check that the total number of inodes is greater than zero before doing
6427 the test of how many are available.
6429 58. When a "warn" ACL statement has a log_message modifier, the message is
6430 remembered, and not repeated. This is to avoid a lot of repetition when a
6431 message has many recipients that cause the same warning to be written.
6432 However, Exim was preserving the list of already written lines for an
6433 entire SMTP session, which doesn't seem right. The memory is now reset if a
6434 new message is started.
6436 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6437 debugging output unless log_rewrite was also set.
6439 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6440 of (int)(handle) when we know that handle contains (void *)(-1).
6442 61. The Exim daemon panic-logs an error return when it closes the incoming
6443 connection. However "connection reset by peer" seems to be common, and
6444 isn't really an error worthy of noting specially, so that particular error
6447 62. When Exim is trying to find all the local interfaces, it used to panic and
6448 die if the ioctl to get the interface flags failed. However, it seems that
6449 on at least one OS (Solaris 9) it is possible to have an interface that is
6450 included in the list of interfaces, but for which you get a failure error
6451 for this call. This happens when the interface is not "plumbed" into a
6452 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6453 failure of the "get flags" call assumes that the interface is down.
6455 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6456 makes life easier for people who are doing arithmetic on fields extracted
6457 from dates, where you often get leading zeros that should not be
6458 interpreted as octal.
6460 64. Added qualify_domain to the redirect router, to override the global
6463 65. If a pathologically long header line contained very many addresses (the
6464 report of this problem mentioned 10 000) and each of them was rewritten,
6465 Exim could use up a very large amount of memory. (It kept on making new
6466 copies of the header line as it rewrote, and never released the old ones.)
6467 At the expense of a bit more processing, the header rewriting function has
6468 been changed so that it no longer eats memory in this way.
6470 66. The generation of the Received: header has been moved from the time that a
6471 message starts to be received, to the time that it finishes. The timestamp
6472 in the Received: header should now be very close to that of the <= log
6473 line. There are two side-effects of this change:
6475 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6476 logged header lines no longer include the local Received: line, because
6477 it has not yet been created. The same applies to a copy of the message
6478 that is returned to a non-SMTP sender when a message is rejected.
6480 (b) When a filter file is tested using -bf, no additional Received: header
6481 is added to the test message. After some thought, I decided that this
6484 This change does not affect the value of $received_for. It is still set
6485 after address rewriting, but before local_scan() is called.
6487 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6489 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6490 gave an unhelpful panic error message, and a defer error. I have managed to
6491 change this behaviour so that it now rejects any supplied certificate,
6492 which seems right, as the list of acceptable certificates is empty.
6494 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6495 gave an unhelpful defer error. I have not managed to make this reject any
6496 supplied certificates, but the error message it gives is "no certificate
6497 supplied", which is not helpful.
6499 70. exigrep's output now also includes lines that are not associated with any
6500 message, but which match the given pattern. Implemented by a patch from
6501 Martin Sluka, which also tidied up the Perl a bit.
6503 71. Recipient callout verification, like sender verification, was using <> in
6504 the MAIL FROM command. This isn't really the right thing, since the actual
6505 sender may affect whether the remote host accepts the recipient or not. I
6506 have changed it to use the actual sender in the callout; this means that
6507 the cache record is now keyed on a recipient/sender pair, not just the
6508 recipient address. There doesn't seem to be a real danger of callout loops,
6509 since a callout by the remote host to check the sender would use <>.
6510 [SEE ABOVE: changed after hitting problems.]
6512 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6513 temporary errors. However, in the case of such a code being given after
6514 the end of a data transmission (i.e. after ".") Exim was failing to write
6515 a retry record for the message. (Yes, there was some broken host that was
6516 actually sending 8xx at this point.)
6518 73. An unknown lookup type in a host list could cause Exim to panic-die when
6519 the list was checked. (An example that provoked this was putting <; in the
6520 middle of a list instead of at the start.) If this happened during a DATA
6521 ACL check, a -D file could be left lying around. This kind of configuration
6522 error no longer causes Exim to die; instead it causes a defer error. The
6523 incident is still logged to the main and panic logs.
6525 74. Buglet left over from Exim 3 conversion. The message "too many messages
6526 in one connection" was written to the rejectlog but not the mainlog, except
6527 when address rewriting (yes!) was being logged.
6529 75. Added write_rejectlog option.
6531 76. When a system filter was run not as root (that is, when system_filter_user
6532 was set), the values of the $n variables were not being returned to the
6533 main process; thus, they were not subsequently available in the $sn
6536 77. Added +return_path_on_delivery log selector.
6538 78. A connection timeout was being treated differently from recipients deferred
6539 when testing hosts_max_try with a message that was older than the host's
6540 retry timeout. (The host should not be counted, thus allowing all hosts to
6541 be tried at least once before bouncing.) This may have been the cause of an
6542 occasionally reported bug whereby a message would remain on the queue
6543 longer than the retry timeout, but would be bounced if a delivery was
6544 forced. I say "may" because I never totally pinned down the problem;
6545 setting up timeout/retry tests is difficult. See also the next item.
6547 79. The ultimate address timeout was not being applied to errors that involved
6548 a combination of host plus message (for example, a timeout on a MAIL
6549 command). When an address resolved to a number of possible hosts, and they
6550 were not all tried for each delivery (e.g. because of hosts_max_try), a
6551 message could remain on the queue longer than the retry timeout.
6553 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6556 81. Fixed an obscure SMTP outgoing bug which required at least the following
6557 conditions: (a) there was another message waiting for the same server;
6558 (b) the server returned 5xx to all RCPT commands in the first message so
6559 that the message was not completed; (c) the server dropped the connection
6560 or gave a negative response to the RSET that Exim sends to abort the
6561 transaction. The observed case was a dropped connection after DATA that had
6562 been sent in pipelining mode. That is, the server had advertised PIPELINING
6563 but was not implementing it correctly. The effect of the bug was incorrect
6564 behaviour, such as trying another host, and this could lead to a crash.
6570 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6571 and daemon.c were passed as pointers to ints; they should have been
6572 pointers to socklen_t variables (which are typically unsigned ints).
6574 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6577 3. Fixed a really odd bug that affected only the testing scheme; patching a
6578 certain fixed string in the binary changed the value of another string that
6579 happened to be identical to the end of the original first string.
6581 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6582 name", it returns that address as the IP address. On some operating
6583 systems (e.g. Solaris), it also passes back the IP address string as the
6584 "host name". However, on others (e.g. Linux), it passes back an empty
6585 string. Exim wasn't checking for this, and was changing the host name to an
6586 empty string, assuming it had been canonicalized.
6588 5. Although rare, it is permitted to have more than one PTR record for a given
6589 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6590 all the names associated with an address, because they do in Solaris.
6591 However, it seems that they do not in Linux for data that comes from the
6592 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6593 I found this out when I moved to a new Linux workstation and tried to run
6594 the Exim test suite.
6596 To get round this problem I have changed the code so that it now does its
6597 own call to the DNS to look up PTR records when searching for a host name.
6598 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6599 addresses that are only in /etc/hosts are still found.
6601 This behaviour is, however, controlled by an option called host_lookup_
6602 order, which defaults to "bydns:byaddr". If people want to use the other
6603 order, or indeed, just use one or the other means of lookup, they can
6604 specify it in this variable.
6606 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6607 some operating systems, this comes back from gethostbyaddr() as an empty
6608 string, and this is what Exim used to test for. However, it seems that in
6609 other systems, "." is yielded. Exim now tests for this case too.
6611 7. The values of check_spool_space and check_log_space are now held internally
6612 as a number of kilobytes instead of an absolute number of bytes. If a
6613 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6614 kilobyte. This means that much larger values can be stored.
6616 8. Exim monitor: an attempt to get the action menu when not actually pointing
6617 at a message produces an empty menu entitled "No message selected". This
6618 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6619 no entries in it ("Shell widget menu has zero width and/or height"). So I
6620 have added a single, blank menu entry in this case.
6622 9. Added ${quote_local_part.
6624 10. MIME decoding is now applied to the contents of Subject: header lines when
6627 11. Now that a reference to $sender_host_address automatically causes a reverse
6628 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6629 host lookup before query-style lookups in lists that might use this
6630 variable. This has therefore been abolished, and the "net-" prefix is no
6631 longer necessary for query-style lookups.
6633 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6634 have been a typo for LFLAGS, so it has been changed.
6636 13. The install script calls Exim with "-C /dev/null" in order to find the
6637 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6638 to be output. However, since Exim outputs its version number before the
6639 error, it didn't break the script. It just looked ugly. I fixed this by
6640 always allowing "-C /dev/null" if the caller is root.
6642 14. Ignore overlarge ACL variable number when reading spool file - insurance
6643 against a later release with more variables having written the file.
6645 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6646 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6647 this, as well as the form without the "IPv6" on the front.
6649 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6650 OS/Makefile-Darwin file.
6652 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6653 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6655 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6656 variable. However, some OS use ints for inodes. I've added cast to long int
6657 to get rid of the compiler warning.
6659 19. I had forgotten to lock out "/../" in configuration file names when
6660 ALT_CONFIG_PREFIX was set.
6662 20. Routers used for verification do not need to specify transports. However,
6663 if such a router generated a host list, and callout was configured, Exim
6664 crashed, because it could not find a port number from the (non-existent)
6665 transport. It now assumes port 25 in this circumstance.
6667 21. Added the -t option to exigrep.
6669 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6670 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6671 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6672 them is set, LOOKUP_LSEARCH is forced.
6674 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6676 24. Added sender and host information to the "rejected by local_scan()" log
6677 line; previously there was no indication of these.
6679 25. Added .include_if_exists.
6681 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6682 Linux. It turns out that not all file systems support this. Apparently some
6683 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6684 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6685 not supported on the file descriptor, is now ignored when Exim is trying to
6686 sync a directory. This applies only to Linux.
6688 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6690 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6691 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6694 29. Added support for maildirsize files from supplied patch (modified a bit).
6696 30. The use of :fail: followed by an empty string could lead Exim to respond to
6697 sender verification failures with (e.g.):
6699 550 Verification failed for <xxx>
6700 550 Sender verify failed
6702 where the first response line was missing the '-' that indicates it is not
6703 the final line of the response.
6705 31. The loop for finding the name of the user that called Exim had a hardwired
6706 limit of 10; it now uses the value of finduser_retries, which is used for
6707 all other user lookups.
6709 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6712 33. Exim was neglecting to zero errno before one call of strtol() when
6713 expanding a string and expecting an integer value. On some systems this
6714 resulted in spurious "integer overflow" errors. Also, it was casting the
6715 result into an int without checking.
6717 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6718 did not work. The code looks as if it has *never* worked, though it appears
6719 to have been documented since at least release 1.62. I have made it work.
6721 35. The "timeout_DNS" error in retry rules, also documented since at least
6722 1.62, also never worked. As it isn't clear exactly what this means, and
6723 clearly it isn't a major issue, I have abolished the feature by treating it
6724 as "timeout", and writing a warning to the main and panic logs.
6726 36. The display of retry rules for -brt wasn't always showing the error code
6729 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6730 timeout_connect_A, timeout_connect_MX.
6732 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6733 to the empty sender.
6735 39. The daemon was not analysing the content of -oX till after it had closed
6736 stderr and disconnected from the controlling terminal. This meant that any
6737 syntax errors were only noted on the panic log, and the return code from
6738 the command was 0. By re-arranging the code a little, I've made the
6739 decoding happen first, so such errors now appear on stderr, and the return
6740 code is 1. However, the actual setting up of the sockets still happens in
6741 the disconnected process, so errors there are still only recorded on the
6744 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6745 connections (as happens on some IP stacks) was logged at start up time as
6746 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6747 it from "IPv6 and IPv4", which means that two separate sockets are being
6750 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6751 says whether AF_INET or AF_INET6 was passed as an argument.
6753 42. Exiwhat output was messed up when time zones were included in log
6756 43. Exiwhat now gives more information about the daemon's listening ports,
6757 and whether -tls-on-connect was used.
6759 44. The "port" option of the smtp transport is now expanded.
6761 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6762 silently ignored. Now an error message is written to the main and panic
6765 46. There's a new ACL modifier called "logwrite" which writes to a log file
6766 as soon as it is encountered.
6768 47. Added $local_user_uid and $local_user_gid at routing time.
6770 48. Exim crashed when trying to verify a sender address that was being
6773 49. Exim was recognizing only a space character after ".include". It now also
6774 recognizes a tab character.
6776 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6777 extracting the relevant information from the specification. The man page no
6778 longer contains scrambled data for the -d option, and I've added a section
6779 at the front about calling Exim under different names.
6781 51. Added "extra_headers" argument to the "mail" command in filter files.
6783 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6786 53. Installed eximstats 1.29.
6788 54. Added transport_filter_timeout as a generic transport option.
6790 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6791 Cc: header lines. This was required by RFC 822, but it not required by RFC
6794 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6795 incoming messages that did not have them. Now it does so only if the
6796 message originates locally, that is, if there is no associated remote host
6797 address. When Resent- header lines are present, this applies to the Resent-
6798 lines rather than the non-Resent- lines.
6800 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6801 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6803 58. Messages for configuration errors now include the name of the main
6804 configuration file - useful now that there may be more than one file in a
6805 list (.included file names were always shown).
6807 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6808 for those rare installations that do not start the daemon as root or run it
6809 setuid root. I've cut out the call to initgroups() if the daemon is not
6812 60. The Exim user and group can now be bound into the binary as text strings
6813 that are looked up at the start of Exim's processing.
6815 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6817 62. Added $mailstore_basename variable.
6819 63. Installed patch to sieve.c from Michael Haardt.
6821 64. When Exim failed to open the panic log after failing to open the main log,
6822 the original message it was trying to log was written to stderr and debug
6823 output, but if they were not available (the usual case in production), it
6824 was lost. Now it is written to syslog before the two lines that record the
6825 failures to open the logs.
6827 65. Users' Exim filters run in subprocesses under the user's uid. It is
6828 possible for a "deliver" command or an alias in a "personal" command to
6829 provoke an address rewrite. If logging of address rewriting is configured,
6830 this fails because the process is not running as root or exim. There may be
6831 a better way of dealing with this, but for the moment (because 4.30 needs
6832 to be released), I have disabled address rewrite logging when running a
6833 filter in a non-root, non-exim process.
6839 1. The buildconfig auxiliary program wasn't quoting the value set for
6840 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6841 not defined. This bug was masked in 4.22 by the effect that was fixed in
6844 2. Some messages that were rejected after a message id was allocated were
6845 shown as "incomplete" by exigrep. It no longer does this for messages that
6846 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6848 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6849 have allow_domain_literals set, the ID did not get logged in the <= line.
6850 Domain literals are now always recognized in Message-ID: header lines.
6852 4. The first argument for a ${extract expansion item is the key name or field
6853 number. Leading and trailing spaces in this item were not being ignored,
6854 causing some misleading effects.
6856 5. When deliver_drop_privilege was set, single queue runner processes started
6857 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6858 same command in the process it spins off) were not dropping privilege.
6860 6. When the daemon running as "exim" started a queue runner, it always
6861 re-executed Exim in the spun-off process. This is a waste of effort when
6862 deliver_drop_privilege is set. The new process now just calls the
6863 queue-runner function directly.
6869 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6872 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6873 ignored. Though the use of -oP was forcing the writing of a pid file, it
6874 was always written to the default place.
6876 3. If the message "no IP address found for host xxxx" is generated during
6877 incoming verification, it is now followed by identification of the incoming
6878 connection (so you can more easily find what provoked it).
6880 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6882 5. Added some features to "harden" Exim a bit more against certain attacks:
6884 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6885 be put in Local/Makefile. This is like the never_users runtime option,
6886 but it cannot be overridden. The default setting is "root".
6888 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6889 prefix string with which any file named in a -C command line option
6892 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6893 is retained for -C and -D only if the caller of Exim is root. Without
6894 it, the exim user may also use -C and -D and retain privilege.
6896 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6897 command line option is disabled.
6899 6. Macro names set by the -D option must start with an upper case letter, just
6900 like macro names defined in the configuration file.
6902 7. Added "dereference=" facility to LDAP.
6904 8. Two instances of the typo "uknown" in the source files are fixed.
6906 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6907 the Configure-Makefile script screwed up while processing it.
6909 10. Incorporated PCRE 4.4.
6911 11. The SMTP synchronization check was not operating right at the start of an
6912 SMTP session. For example, it could not catch a HELO sent before the client
6913 waited for the greeting. There is now a check for outstanding input at the
6914 point when the greeting is written. Because of the duplex, asynchronous
6915 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6916 way, but not yet received, when the check is performed.
6918 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6919 on TCP/IP sockets, because this apparently causes some broken clients to
6922 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6923 unchanged) from the Cygwin maintainer.
6925 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6926 was supported instead of testing for mbx. Effectively a typo.
6928 15. The spa authenticator server code was not checking that the input it
6929 received was valid base64.
6931 16. The debug output line for the "set" modifier in ACLs was not showing the
6932 name of the variable that was being set.
6934 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6936 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6937 lookup on its own. Something else was needed to trigger the lookup. For
6938 example, a match in host_lookup or the need for a host name in a host list.
6939 Now, if $sender_host_name is referenced and the host name has not yet been
6940 looked up, a lookup is performed. If the lookup fails, the variable remains
6941 empty, and $host_lookup_failed is set to "1".
6943 19. Added "eqi" as a case-independent comparison operator.
6945 20. The saslauthd authentication condition could segfault if neither service
6946 nor realm was specified.
6948 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6949 error message that was logged was misleading, and incoming SMTP
6950 connections were dropped. The message is now more accurate, and temporary
6951 errors are given to SMTP connections.
6953 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6954 (except RSET) with a 421 error, until QUIT is received. However, it was
6955 failing to send a response to QUIT.
6957 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6958 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6959 ACL is now run after the helo_try_verify_hosts code.
6961 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6962 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6963 case-independent, so other case variants are also recognized). Apparently
6964 some systems use these upper case variants.
6966 25. If more than two messages were waiting for the same host, and a transport
6967 filter was specified for the transport, Exim sent two messages over the
6968 same TCP/IP connection, and then failed with "socket operation on non-
6969 socket" when it tried to send the third.
6971 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6973 27. The extern definition of crypt16() in expand.c was not being excluded when
6974 the OS had its own crypt16() function.
6976 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6977 as a preferred synonym for return_size_limit, both as an option and as an
6980 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6982 30. Changed the default configuration ACL to relax the local part checking rule
6983 for addresses that are not in any local domains. For these addresses,
6984 slashes and pipe symbols are allowed within local parts, but the sequence
6985 /../ is explicitly forbidden.
6987 31. SPA server authentication was not clearing the challenge buffer before
6990 32. log_message in a "warn" ACL statement was writing to the reject log as
6991 well as to the main log, which contradicts the documentation and doesn't
6992 seem right (because no rejection is happening). So I have stopped it.
6994 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6995 However, I am unable to do any testing of this.
6997 34. Fixed an infelicity in the appendfile transport. When checking directories
6998 for a mailbox, to see if any needed to be created, it was accidentally
6999 using path names with one or more superfluous leading slashes; tracing
7000 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7002 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7003 contents are added to the log line that is written for every discarded
7004 recipient. (Previously a log_message setting was ignored.)
7006 36. The ${quote: operator now quotes the string if it is empty.
7008 37. The install script runs exim in order to find its version number. If for
7009 some reason other than non-existence or emptiness, which it checks, it
7010 could not run './exim', it was installing it with an empty version number,
7011 i.e. as "exim-". This error state is now caught, and the installation is
7014 38. An argument was missing from the function that creates an error message
7015 when Exim fails to connect to the socket for saslauthd authentication.
7016 This could cause Exim to crash, or give a corrupted message.
7018 39. Added isip, isip4, and isip6 to ${if conditions.
7020 40. The ACL variables $acl_xx are now saved with the message, and can be
7021 accessed later in routers, transports, and filters.
7023 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7024 strings in the file are not string-expanded.
7026 42. If a MAIL command specified a SIZE value that was too large to fit into an
7027 int variable, the check against message_size_limit failed. Such values are
7028 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7029 day this will have to be increased, but I don't think I want to be around
7030 when emails are that large.
7037 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7038 iconv() is not standard in FreeBSD.
7040 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7041 IPv6 enabled. The observed symptom was a segmentation fault on return from
7042 the function os_common_find_running_interfaces() in src/os.c.
7044 3. In the check_special_case() function in daemon.c I had used "errno" as an
7045 argument name, which causes warnings on some systems. This was basically a
7046 typo, since it was named "eno" in the comments!
7048 4. The code that waits for the clock to tick (at a resolution of some fraction
7049 of a second) so as to ensure message-id uniqueness was always waiting for
7050 at least one whole tick, when it could have waited for less. [This is
7051 almost certainly not relevant at current processor speeds, where it is
7052 unlikely to ever wait at all. But we try to future-proof.]
7054 5. The function that sleeps for a time interval that includes fractions of a
7055 second contained a race. It did not block SIGALRM between setting the
7056 timer, and suspending (a couple of lines later). If the interval was short
7057 and the sigsuspend() was delayed until after it had expired, the suspension
7058 never ended. On busy systems this could lead to processes getting stuck for
7061 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7062 process, before it forks any delivery processes. The open lookup caching
7063 mechanism meant that the open file or database connection was passed into
7064 the delivery process. The problem was that delivery processes always tidy
7065 up cached lookup data. This could cause a problem for the next delivery
7066 process started by the queue runner, because the external queue runner
7067 process does not know about the closure. So the next delivery process
7068 still has data in the lookup cache. In the case of a file lookup, there was
7069 no problem because closing a file descriptor in a subprocess doesn't affect
7070 the parent. However, if the lookup was caching a connection to a database,
7071 the connection was closed, and the second delivery process was likely to
7072 see errors such as "PGSQL: query failed: server closed the connection
7073 unexpectedly". The problem has been fixed by closing all cached lookups
7074 in a queue runner before running a delivery process.
7076 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7077 seem to have the "const" qualifier which it has on other OS. I've
7080 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7081 *of the same type* (client or server) with the same public name that an
7082 error should be diagnosed.
7084 9. When Exim looked up a host name for an IP address, but failed to find the
7085 original IP address when looking up the host name (a safety check), it
7086 output the message "<ip address> does not match any IP for NULL", which was
7087 confusing, to say the least. The bug was that the host name should have
7088 appeared instead of "NULL".
7090 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7091 user that is built into the binary, and the -C or -D options is used, root
7092 privilege is dropped before the configuration file is read. In addition,
7093 logging is switched to stderr instead of the normal log files. If the
7094 configuration then re-defines the Exim user, the unprivileged environment
7095 is probably not what is expected, so Exim logs a panic warning message (but
7098 However, if deliver_drop_privilege is set, the unprivileged state may well
7099 be exactly what is intended, so the warning has been cut out in that case,
7100 and Exim is allowed to try to write to its normal log files.
7106 1. smtp_return_error_details was not giving details for temporary sender
7107 or receiver verification errors.
7109 2. Diagnose a configuration error if two authenticators have the same public
7112 3. Exim used not to create the message log file for a message until the first
7113 delivery attempt. This could be confusing when incoming messages were held
7114 for policy or load reasons. The message log file is now created at the time
7115 the message is received, and an initial "Received" line is written to it.
7117 4. The automatically generated man page for command line options had a minor
7118 bug that caused no ill effects; however, a more serious problem was that
7119 the procedure for building the man page automatically didn't always
7120 operate. Consequently, release 4.20 contains an out-of-date version. This
7121 shouldn't happen again.
7123 5. When building Exim with embedded Perl support, the script that builds the
7124 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7125 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7127 6. The freeze_tell option was not being used for messages that were frozen on
7128 arrival, either by an ACL or by local_scan().
7130 7. Added the smtp_incomplete_transaction log selector.
7132 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7133 was accepting AUTH without a new EHLO.
7135 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7136 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7139 10. Logging of TCP/IP connections (when configured) now happens in the main
7140 daemon process instead of the child process, so that the TCP/IP connection
7141 count is more accurate (but it can never be perfect).
7143 11. The use of "drop" in a nested ACL was not being handled correctly in the
7144 outer ACL. Now, if condition failure induced by the nested "drop" causes
7145 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7146 or "require"), the connection is dropped.
7148 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7149 that yield "discard" can now be used with an "accept" or a "discard" verb,
7150 but an error is generated for any others (because I can't see a useful way
7151 to define what should happen).
7153 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7154 are now processed in the same way as lines in the Exim configuration file.
7155 In particular, continuation lines are supported.
7157 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7159 15. Added -ti meaning -t -i.
7161 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7162 domains, and warn by logging if others are found.
7164 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7165 returned from GIFCONF when Exim is trying to find the list of interfaces on
7166 a host. The code in os.c has been modified to copy each ifreq to an aligned
7167 structure in all cases.
7169 Also, in some cases, the returned ifreq's were being copied to a 'struct
7170 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7171 means the last couple of bytes of an IPv6 address could be chopped if the
7172 ifreq contained only a normal sockaddr (14 bytes storage).
7174 18. Named domain lists were not supported in the hosts_treat_as_local option.
7175 An entry such as +xxxx was not recognized, and was treated as a literal
7178 19. Ensure that header lines added by a DATA ACL are included in the reject log
7179 if the ACL subsequently rejects the message.
7181 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7182 MD5 (which is deprecated).
7184 21. When testing a filter file using -bf, Exim was writing a message when it
7185 took the sender from a "From " line in the message, but it was not doing so
7186 when it took $return_path from a Return-Path: header line. It now does.
7188 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7189 with a valid header line field name (a series of printing characters
7190 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7192 23. Changed "disc" in the source to "disk" to conform to the documentation and
7193 the book and for uniformity.
7195 24. Ignore Sendmail's -Ooption=value command line item.
7197 25. When execve() failed while trying to run a command in a pipe transport,
7198 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7199 could be confused with a return value of 69 from the command itself. This
7200 has been changed to 127, the value the shell returns if it is asked to run
7201 a non-existent command. The wording for the related log line suggests a
7202 non-existent command as the problem.
7204 26. If received_header_text expands to an empty string, do not add a Received:
7205 header line to the message. (Well, it adds a token one on the spool, but
7206 marks it "old" so that it doesn't get used or transmitted.)
7208 27. Installed eximstats 1.28 (addition of -nt option).
7210 28. There was no check for failure on the call to getsockname() in the daemon
7211 code. This can fail if there is a shortage of resources on the system, with
7212 ENOMEM, for example. A temporary error is now given on failure.
7214 29. Contrary to the C standard, it seems that in some environments, the
7215 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7216 program. Exim now does this explicitly; it affects the formatting of
7217 timestamps using strftime().
7219 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7220 complaints. I've now initialized all the variables, to avoid this.
7222 32. Header lines added by a system filter were not being "seen" during
7223 transport-time rewrites.
7225 33. The info_callback() function passed to OpenSSL is set up with type void
7226 (*)(SSL *, int, int), as described somewhere. However, when calling the
7227 function (actually a macro) that sets it up, the type void(*)() is
7228 expected. I've put in a cast to prevent warnings from picky compilers.
7230 34. If a DNS black list lookup found a CNAME record, but there were no A
7231 records associated with the domain it pointed at, Exim crashed.
7233 35. If a DNS black list lookup returned more than one A record, Exim ignored
7234 all but the first. It now scans all returned addresses if a particular IP
7235 value is being sought. In this situation, the contents of the
7236 $dnslist_value variable are a list of all the addresses, separated by a
7239 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7240 to accept a host name and all its aliases if the forward lookup for any of
7241 them yielded the IP address of the incoming connection. Now it accepts only
7242 those names whose forward lookup yields the correct IP address. Any other
7243 names are discarded. This closes a loophole whereby a rogue DNS
7244 administrator could create reverse DNS records to break through a
7245 wildcarded host restriction in an ACL.
7247 37. If a user filter or a system filter that ran in a subprocess used any of
7248 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7249 the wrong values were passed to the pipe command ($thisaddress had the
7250 value of $0, $0 had the value of $1, etc). This bug was introduced by
7251 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7253 38. Improved the line breaking for long SMTP error messages from ACLs.
7254 Previously, if there was no break point between 40 and 75 characters, Exim
7255 left the rest of the message alone. Two changes have been made: (a) I've
7256 reduced the minimum length to 35 characters; (b) if it can't find a break
7257 point between 35 and 75 characters, it looks ahead and uses the first one
7258 that it finds. This may give the occasional overlong line, but at least the
7259 remaining text gets split now.
7261 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7262 file descriptors might be low, and that setting 1000 would always raise it.
7263 It turns out that in some environments, the limit is already over 1000 and
7264 that lowering it causes trouble. So now Exim takes care not to decrease it.
7266 40. When delivering a message, the value of $return_path is set to $sender_
7267 address at the start of routing (routers may change the value). By an
7268 oversight, this default was not being set up when an address was tested by
7269 -bt or -bv, which affected the outcome if any router or filter referred to
7272 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7273 log only when "message" or "log_message" are set. However, if one of the
7274 conditions was an address verification, or a call to a nested ACL, the
7275 messages generated by the underlying test were being passed through. This
7276 no longer happens. The underlying message is available in $acl_verify_
7277 message for both "message" and "log_message" expansions, so it can be
7278 passed through if needed.
7280 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7281 new expansion $bh_ to give the encoded byte string without charset
7282 translation. Translation happens only if iconv() is available; HAVE_ICONV
7283 indicates this at build time. HEADERS_CHARSET gives the charset to
7284 translate to; headers_charset can change it in the configuration, and
7285 "headers charset" can change it in an individual filter file.
7287 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7288 that creates RFC 2047 encoded "words" labels them as that charset instead
7289 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7290 expansion operator; (ii) when Exim creates a From: line for a local
7291 message; (iii) when a header line is rewritten to include a "phrase" part.
7293 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7294 buggy, causing it to skip the first lines of messages whose message ID
7295 ended in 'D'. This would not have bitten before Exim release 4.14, because
7296 message IDs were unlikely to end in 'D' before then. The effect was to have
7297 incorrect size information for certain domains.
7299 45. #include "config.h" was missing at the start of the crypt16.c module. This
7300 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7303 46. If there was a timeout during a "random" callout check, Exim treated it as
7304 a failure of the random address, and carried on sending RSET and the real
7305 address. If the delay was just some slowness somewhere, the response to the
7306 original RCPT would be taken as a response to RSET and so on, causing
7307 mayhem of various kinds.
7309 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7310 when I implemented it. It didn't allow for the fact that some option values
7311 may legitimately be negative (e.g. size_addition), and it didn't even do
7312 the right test for positive values.
7314 48. Domain names in DNS records are case-independent. Exim always looks them up
7315 in lower case. Some resolvers return domain names in exactly the case they
7316 appear in the zone file, that is, they may contain uppercase letters. Not
7317 all resolvers do this - some return always lower case. Exim was treating a
7318 change of case by a resolver as a change of domain, similar to a widening
7319 of a domain abbreviation. This triggered its re-routing code and so it was
7320 trying to route what was effectively the same domain again. This normally
7321 caused routing to fail (because the router wouldn't handle the domain
7322 twice). Now Exim checks for this case specially, and just changes the
7323 casing of the domain that it ultimately uses when it transmits the message
7326 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7329 50. If a filter generated a file delivery with a non-absolute name (possible if
7330 no home directory exists for the router), the forbid_file option was not
7333 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7334 the existing equality matching.
7336 52. Exim was using ints instead of ino_t variables in some places where it was
7337 dealing with inode numbers.
7339 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7340 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7341 called TMPDIR, and if it finds it is different, it changes its value.
7343 54. The smtp_printf() function is now made available to local_scan() so
7344 additional output lines can be written before returning. There is also an
7345 smtp_fflush() function to enable the detection of a dropped connection.
7346 The variables smtp_input and smtp_batched_input are exported to
7349 55. Changed the default runtime configuration: the message "Unknown user"
7350 has been removed from the ACL, and instead placed on the localuser router,
7351 using the cannot_route_message feature. This means that any verification
7352 failures that generate their own messages won't get overridden. Similarly,
7353 the "Unrouteable address" message that was in the ACL for unverifiable
7354 relay addresses has also been removed.
7356 56. Added hosts_avoid_esmtp to the smtp transport.
7358 57. The exicyclog script was not checking for the esoteric option
7359 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7360 will work only if exicyclog is run under the appropriate euid.
7362 58. Following a discussion on the list, the rules by which Exim recognises line
7363 endings on incoming messages have been changed. The -dropcr and drop_cr
7364 options are now no-ops, retained only for backwards compatibility. The
7365 following line terminators are recognized: LF CRLF CR. However, special
7366 processing applies to CR:
7368 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7369 nor a local message in the state where . is a terminator.
7371 (ii) If a bare CR is encountered in a header line, an extra space is added
7372 after the line terminator so as not to end the header. The reasoning
7373 behind this is that bare CRs in header lines are most likely either
7374 to be mistakes, or people trying to play silly games.
7376 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7377 was being incorrectly given as 18 bytes larger than it should have been.
7378 This is a VOB (very old bug).
7380 60. This may never have affected anything current, but just in case it has:
7381 When the local host is found other than at the start of a list of hosts,
7382 the local host, those with the same MX, and any that follow, are discarded.
7383 When the list in question was part of a longer list of hosts, the following
7384 hosts (not currently being processed) were also being discarded. This no
7385 longer happens. I'm not sure if this situation could ever has previously
7388 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7391 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7392 header. This is something that is recommended in a new Internet Draft, and
7393 is something that is documented as being done by Sendmail. There are two
7394 possible values. For messages generated by the autoreply transport, Exim
7397 Auto-Submitted: auto-replied
7399 whereas for all other generated messages (e.g. bounces) it adds
7401 Auto-Submitted: auto-generated
7403 63. The "personal" condition in filters now includes a test for the
7404 Auto-Submitted: header. If it contains the string "auto-" the message it
7405 not considered personal.
7407 64. Added rcpt_include_affixes as a generic transport option.
7409 65. Added queue_only_override (default true).
7411 66. Added the syslog_duplication option.
7413 67. If what should have been the first header line of a message consisted of
7414 a space followed by a colon, Exim was mis-interpreting it as a header line.
7415 It isn't of course - it is syntactically invalid and should therefore be
7416 treated as the start of the message body. The misbehaviour could have
7417 caused a number of strange effects, including loss of data in subsequent
7418 header lines, and spool format errors.
7420 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7421 client host had authenticated. This control can now be exercised by an ACL
7422 for more flexibility.
7424 69. By default, callouts do not happen when testing with -bh. There is now a
7425 variant, -bhc, which does actually run the callout code, including
7426 consulting and updating the callout cache.
7428 70. Added support for saslauthd authentication, courtesy of Alexander
7431 71. If statvfs() failed on the spool or log directories while checking their
7432 size for availability, Exim confusingly gave the error "space shortage".
7433 Furthermore, in debugging mode it crashed with a floating point exception.
7434 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7435 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7436 serious problem, Exim now writes to the main and panic logs when this
7437 happens, with details of the failure. It then refuses to accept the
7438 incoming message, giving the message "spool directory problem" or "log
7439 directory problem" with a 421 code for SMTP messages.
7441 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7442 0, 1, and 2 exist, because some OS complain for execs without them (see
7443 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7444 descriptors. However, the code omitted to check that the open succeeded,
7445 causing mysterious errors if for some reason the permissions on /dev/null
7446 got screwed. Now Exim writes a message to the main and panic logs, and
7447 bombs out if it can't open /dev/null.
7449 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7450 interact so that it is all more flexible. It is supposed to remain
7451 backwards compatible. Also added extra_local_interfaces.
7453 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7454 to bomb out with an assertion failure - to the client this appears as a
7455 connection drop. This problem occurs in the part of the code that was taken
7456 from the Samba project. Fortunately, the assertion is in a very simple
7457 function, so I have fixed this by reproducing the function inline in the
7458 one place where it is called, and arranging for authentication to fail
7459 instead of killing the process with assert().
7461 75. The SPA client code was not working when the server requested OEM rather
7462 than Unicode encoding.
7464 76. Added code to make require_files with a specific uid setting more usable in
7465 the case where statting the file as root fails - usually a non-root-mounted
7466 NFS file system. When this happens and the failure is EACCES, Exim now
7467 forks a subprocess and does the per-uid checking as the relevant uid.
7469 77. Added process_log_path.
7471 78. If log_file_path was not explicitly set, a setting of check_log_space or
7472 check_log_inodes was ignored.
7474 79. If a space check for the spool or log partitions fails, the incident is now
7475 logged. Of course, in the latter case the data may get lost...
7477 80. Added the %p formatting code to string_format() so that it can be used to
7478 print addresses in debug_print(). Adjusted all the address printing in the
7479 debugging in store.c to use %p rather than %d.
7481 81. There was a concern that a line of code in smtp_in.c could overflow a
7482 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7483 initially expressed, the concern was not well-founded, because trailing
7484 spaces are removed early. However, if the trailing spaces were followed by
7485 a NULL, they did not get removed, so the overflow was possible. Two fixes
7488 (a) I re-wrote the offending code in a cleaner fashion.
7489 (b) If an incoming SMTP command contains a NULL character, it is rejected
7492 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7493 runs initgroups(), so that if the Exim user is in any additional groups,
7494 they will be used during message reception.
7500 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob