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.
136 JH/01 Remove code calling the customisable local_scan function, unless a new
137 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
139 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
140 non-signal-safe functions being used.
142 JH/03 Bug 2269: When presented with a received message having a stupidly large
143 number of DKIM-Signature headers, disable DKIM verification to avoid
144 a resource-consumption attack. The limit is set at twenty.
146 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
147 report of oldest_pass in ${authres } in consequence, and separate out
148 some descriptions of reasons for verification fail.
150 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
151 files in the spool were present and unlocked. A queue-runner could spot
152 them, resulting in a duplicate delivery. Fix that by doing the unlock
153 after the unlink. Investigation by Tim Stewart. Take the opportunity to
154 add more error-checking on spoolfile handling while that code is being
157 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
158 No known attacks, no CVE, this is defensive hardening.
160 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
161 a queue-runner could start a delivery while other operations were ongoing.
162 Cutthrough delivery was a common victim, resulting in duplicate delivery.
163 Found and investigated by Tim Stewart. Fix by using the open message data
164 file handle rather than opening another, and not locally closing it (which
165 releases a lock) for that case, while creating the temporary .eml format
166 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
168 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
169 $sender_verify_failure/$recipient_verify_failure to "random".
171 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
174 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
175 Previously this would segfault.
177 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
180 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
181 like zero, since the resolver should be doing this for us, But we need one
182 as a CNAME but no MX presence gets the CNAME returned; we need to check
183 that doesn't point to an MX to declare it "no MX returned" rather than
184 "error, loop". A new main option is added so the older capability of
185 following some limited number of chain links is maintained.
187 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
189 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
190 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
191 now-deprecated earlier definitions used only specified the range up to TLS
192 1.2 (in the older-version library docs).
194 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
196 JH/15 Rework TLS client-side context management. Stop using a global, and
197 explicitly pass a context around. This enables future use of TLS for
198 connections to service-daemons (eg. malware scanning) while a client smtp
199 connection is using TLS; with cutthrough connections this is quite likely.
201 JH/16 Fix ARC verification to do AS checks in reverse order.
203 JH/17 Support a "tls" option on the ${readsocket } expansion item.
205 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
206 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
207 Previously the "utf8" would be re-prepended for every additional message.
209 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
210 Previously thery were accepted, resulting in issues when attempting to
211 forward messages to a non-supporting MTA.
213 PP/02 Let -n work with printing macros too, not just options.
215 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
216 one parent address was copied, and bogus data was used at delivery-logging
217 time. Either a crash (after delivery) or bogus log data could result.
218 Discovery and analysis by Tim Stewart.
220 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
221 Previously if the string ended mid-character, we did not insert the
222 promised '?' replacement.
224 PP/04 Documentation: current string operators work on bytes, not codepoints.
226 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
227 should pack well giving a smaller memory footprint so better caching and
228 therefore performance. Group the declarations where this can't be done so
229 that the byte-sized flag variables are not interspersed among pointer
230 variables, giving a better chance of good packing by the compiler.
232 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
233 non-null, to avoid issues with sites running BATV. Previously reports were
234 sent with an empty envelope sender so looked like bounces.
236 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
237 The ignore_error flag wasn't being returned from the filter subprocess so
238 was not set for later routers. Investigation and fix by Matthias Kurz.
240 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
241 and a msg:complete for the whole, when a message is manually removed using
242 -Mrm. Developement by Matthias Kurz, hacked on by JH.
244 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
245 a "Gnu special" function, asprintf() in the DB utility binary builds; I
246 hope that is portable enough.
248 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
249 requiring a known-CA anchor certificate; make it now rely entirely on the
250 TLSA as an anchor. Checking the name on the leaf cert against the name
251 on the A-record for the host is still done for TA (but not for EE mode).
253 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
254 included in delivery lines for non-proxied connections, when compiled with
255 SUPPORT_SOCKS and running with proxy logging enabled.
257 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
258 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
259 move the existing event to fire before the normal logging of message
260 failure so that custom logging is bracketed by normal logging.
262 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
263 msg:fail:internal event. Developement by Matthias Kurz.
265 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
266 far too small for todays use of crypto signatures stored there. Go all
267 the way to the max DNS message size of 64kB, even though this might be
268 overmuch for IOT constrained device use.
270 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
271 copy a string over itself. The library routine is documented as not
272 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
274 JH/32 For main options check_spool_space and check_inode_space, where the
275 platform supports 64b integers, support more than the previous 2^31 kB
276 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
277 the previous G, M, k.
279 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
280 $authenticated_fail_id variable on authentication failure. Previously
283 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
284 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
285 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
286 GNUTLS_SEC_PARAM_MEDIUM.
288 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
289 side. Previously we would continue as if no SNI had been received.
291 JH/36 Harden the handling of string-lists. When a list consisted of a sole
292 "<" character, which should be a list-separator specification, we walked
293 off past the nul-terimation.
295 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
296 causes) even when the retry time is not yet met. Previously they were
297 not, meaning that when (say) an account was over-quota and temp-rejecting,
298 and multiple senders' messages were queued, only one sender would get
299 notified on each configured delay_warning cycle.
301 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
303 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
304 cipher-suites, an error can be left on the stack even for a succeeding
305 accept; this results in impossible error messages when a later operation
308 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
309 return error codes indicating retry. Under TLS1.3 this becomes required.
311 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
312 it only wrote the new authenticators, resulting in a lack of tracking of
313 peer changes of ESMTP extensions until the next cache flush.
315 JH/41 Fix the loop reading a message header line to check for integer overflow,
316 and more-often against header_maxsize. Previously a crafted message could
317 induce a crash of the recive process; now the message is cleanly rejected.
319 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
320 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
326 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
327 When redis_servers is set to a list of > 1 element, and the Redis servers
328 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
329 case of MOVED into a DEFER case instead, thus moving the query onto the
330 next server in the list. For a cluster of N elements, all N servers must
331 be defined in redis_servers.
333 GF/02 Catch and remove uninitialized value warning in exiqsumm
334 Check for existence of @ARGV before looking at $ARGV[0]
336 JH/01 Replace the store_release() internal interface with store_newblock(),
337 which internalises the check required to safely use the old one, plus
338 the allocate and data copy operations duplicated in both (!) of the
339 extant use locations.
341 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
342 modifier. This matches the restriction on the commandline.
344 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
345 Previously only the last row was returned.
347 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
348 we assumed that tags in the header were well-formed, and parsed the
349 element content after inspecting only the first char of the tag.
350 Assumptions at that stage could crash the receive process on malformed
353 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
354 While running the DKIM ACL we operate on the Permanent memory pool so that
355 variables created with "set" persist to the DATA ACL. Also (at any time)
356 DNS lookups that fail create cache records using the Permanent pool. But
357 expansions release any allocations made on the current pool - so a dnsdb
358 lookup expansion done in the DKIM ACL releases the memory used for the
359 DNS negative-cache, and bad things result. Solution is to switch to the
360 Main pool for expansions.
361 While we're in that code, add checks on the DNS cache during store_reset,
362 active in the testsuite.
363 Problem spotted, and debugging aided, by Wolfgang Breyha.
365 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
366 When none of the hosts presented to a transport match an already-open
367 connection, close it and proceed with the list. Previously we would
368 queue the message. Spotted by Lena with Yahoo, probably involving
371 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
372 Previously a spurious "250 OK id=" response was appended to the proper
375 JH/08 The "support for" informational output now, which built with Content
376 Scanning support, has a line for the malware scanner interfaces compiled
377 in. Interface can be individually included or not at build time.
379 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
380 by the template makefile "src/EDITME". The "STREAM" support for an older
381 ClamAV interface method is removed.
383 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
384 rows affected is given instead).
386 JH/11 The runtime Berkeley DB library version is now additionally output by
387 "exim -d -bV". Previously only the compile-time version was shown.
389 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
390 SMTP connection. Previously, when one had more receipients than the
391 first, an abortive onward connection was made. Move to full support for
392 multiple onward connections in sequence, handling cutthrough connection
393 for all multi-message initiating connections.
395 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
396 routers. Previously, a multi-recipient message would fail to match the
397 onward-connection opened for the first recipient, and cause its closure.
399 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
400 a timeout on read on a GnuTLS initiating connection, resulting in the
401 initiating connection being dropped. This mattered most when the callout
402 was marked defer_ok. Fix to keep the two timeout-detection methods
405 JH/15 Relax results from ACL control request to enable cutthrough, in
406 unsupported situations, from error to silently (except under debug)
407 ignoring. This covers use with PRDR, frozen messages, queue-only and
410 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
412 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
413 metadata, resulting in a crash in free().
415 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
416 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
417 Broken also in d185889f4, with init system revamp.
419 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
420 Previously we abruptly closed the connection after reading a malware-
421 found indication; now we go on to read the "scan ok" response line,
424 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
425 ACL. Previously, a crash would result.
427 JH/19 Speed up macro lookups during configuration file read, by skipping non-
428 macro text after a replacement (previously it was only once per line) and
429 by skipping builtin macros when searching for an uppercase lead character.
431 JH/20 DANE support moved from Experimental to mainline. The Makefile control
432 for the build is renamed.
434 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
435 was allocated for every new TLS startup, meaning one per message. Fix
436 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
438 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
439 reported the original. Fix to report (as far as possible) the ACL
440 result replacing the original.
442 JH/23 Fix memory leak during multi-message connections using STARTTLS under
443 OpenSSL. Certificate information is loaded for every new TLS startup,
444 and the resources needed to be freed.
446 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
448 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
451 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
452 DATA response info to the (existing) per-recipient response info for
453 the "C=" log element. It can have useful tracking info from the
454 destination system. Patch from Simon Arlott.
456 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
457 length value. Previously this would segfault.
459 HS/02 Support Avast multiline protoocol, this allows passing flags to
460 newer versions of the scanner.
462 JH/28 Ensure that variables possibly set during message acceptance are marked
463 dead before release of memory in the daemon loop. This stops complaints
464 about them when the debug_store option is enabled. Discovered specifically
465 for sender_rate_period, but applies to a whole set of variables.
466 Do the same for the queue-runner and queue-list loops, for variables set
467 from spool message files. Do the same for the SMTP per-message loop, for
468 certain variables indirectly set in ACL operations.
470 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
471 as a multi-recipient message from a mailinglist manager). The coding had
472 an arbitrary cutoff number of characters while checking for more input;
473 enforced by writing a NUL into the buffer. This corrupted long / fast
474 input. The problem was exposed more widely when more pipelineing of SMTP
475 responses was introduced, and one Exim system was feeding another.
476 The symptom is log complaints of SMTP syntax error (NUL chars) on the
477 receiving system, and refused recipients seen by the sending system
478 (propating to people being dropped from mailing lists).
479 Discovered and pinpointed by David Carter.
481 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
482 replaced by the ${authresults } expansion.
484 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
486 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
487 allows proper process termination in container environments.
489 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
490 Previously the "final dot" had a newline after it; ensure it is CR,LF.
492 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
493 and "err_perm", deprecated since 4.83 when the RFC-defined words
494 "temperror" and "permerror" were introduced.
496 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
497 transport-filters or DKIM-signing. The restriction was lost in the
498 consolidation of verify-callout and delivery SMTP handling.
499 Extend the restriction to also cover ARC-signing.
501 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
502 in defer=pass mode supply a 450 to the initiator. Previously the message
505 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
506 tls_require_ciphers is used as before.
508 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
509 "pass_unscanned". Only tmpfails from the scanner are written to
510 the paniclog, as they may require admin intervention (permission
511 denied, license issues). Other scanner errors (like decompression
512 bombs) do not cause a paniclog entry.
514 JH/36 Fix reinitialisation of DKIM logging variable between messages.
515 Previously it was possible to log spurious information in receive log
518 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
519 triggered odd behaviour from Outlook Express clients.
521 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
524 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
525 since the IETF WG has not yet settled on that versus the original
526 "bare" representation.
528 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
529 Previously the millisecond value corrupted the output.
530 Fix also for syslog_pid=no and log_selector +pid, for which the pid
531 corrupted the output.
537 JH/01 Rework error string handling in TLS interface so that the caller in
538 more cases is responsible for logging. This permits library-sourced
539 string to be attached to addresses during delivery, and collapses
540 pairs of long lines into single ones.
542 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
543 during configuration. Wildcards are allowed and expanded.
545 JH/02 Rework error string handling in DKIM to pass more info back to callers.
546 This permits better logging.
548 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
549 do not close it down and have the child transport start it up again on
550 the passed-on TCP connection. Instead, proxy the child (and any
551 subsequent ones) for TLS via a unix-domain socket channel. Logging is
552 affected: the continued delivery log lines do not have any DNSSEC, TLS
553 Certificate or OCSP information. TLS cipher information is still logged.
555 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
556 identical IP addresses on different listening ports. Will also affect
559 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
560 add noisy ifdef guards to special-case this sillyness.
561 Patch from Bernd Kuhls.
563 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
564 than 255 are no longer allowed.
566 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
567 Disable the session-cache too, which might reduce our load. Since we
568 currrectly use a new context for every connection, both as server and
569 client, there is no benefit for these.
570 GnuTLS appears to not support tickets server-side by default (we don't
571 call gnutls_session_ticket_enable_server()) but client side is enabled
572 by default on recent versions (3.1.3 +) unless the PFS priority string
575 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
576 <https://reproducible-builds.org/specs/source-date-epoch/>.
578 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
579 the check for any unsuccessful recipients did not notice the limit, and
580 erroneously found still-pending ones.
582 JH/08 Pipeline CHUNKING command and data together, on kernels that support
583 MSG_MORE. Only in-clear (not on TLS connections).
585 JH/09 Avoid using a temporary file during transport using dkim. Unless a
586 transport-filter is involved we can buffer the headers in memory for
587 creating the signature, and read the spool data file once for the
588 signature and again for transmission.
590 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
591 4.77 as the kernel support then wasn't solid, having issues in 64bit
592 mode. Now, it's been long enough. Add support for FreeBSD also.
594 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
595 case where the routing stage had gathered several addresses to send to
596 a host before calling the transport for the first, we previously failed
597 to close down TLS in the old transport process before passing the TCP
598 connection to the new process. The new one sent a STARTTLS command
599 which naturally failed, giving a failed delivery and bloating the retry
600 database. Investigation and fix prototype from Wolfgang Breyha.
602 JH/12 Fix check on SMTP command input synchronisation. Previously there were
603 false-negatives in the check that the sender had not preempted a response
604 or prompt from Exim (running as a server), due to that code's lack of
605 awareness of the SMTP input buffering.
607 PP/04 Add commandline_checks_require_admin option.
608 Exim drops privileges sanely, various checks such as -be aren't a
609 security problem, as long as you trust local users with access to their
610 own account. When invoked by services which pass untrusted data to
611 Exim, this might be an issue. Set this option in main configuration
612 AND make fixes to the calling application, such as using `--` to stop
615 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
616 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
617 responses to those, into a single TLS record each way (this usually means
618 a single packet). As a side issue, smtp_enforce_sync now works on TLS
621 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
622 affects you only if you're dancing at the edge of the param size limits.
623 If you are, and this message makes sense to you, then: raise the
624 configured limit or use OpenSSL 1.1. Nothing we can do for older
627 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
628 cmdline element to get the documented default one. Previously it was
631 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
632 are unsuable for non-message-specific reasons - eg. connection timeout,
633 banner-time rejection.
635 JH/16 Fix logging of delivery remote port, when specified by router, under
638 PP/06 Repair manualroute's ability to take options in any order, even if one
639 is the name of a transport.
642 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
644 JH/17 Change the list-building routines interface to use the expanding-string
645 triplet model, for better allocation and copying behaviour.
647 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
648 Previously it was constructed the first time a possibly-matching string
649 was met in the configuration file input during startup; now it is done
652 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
653 compatible one, to avoid the (poorly documented) possibility of a config
654 file in the working directory redirecting the DB files, possibly correpting
655 some existing file. CVE-2017-10140 assigned for BDB.
657 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
658 cache-hot. Previously, although the result was properly cached, the
659 initial verify call returned a defer.
661 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
662 the main verify for receipient in uncached-mode.
664 JH/22 Retire historical build files to an "unsupported" subdir. These are
665 defined as "ones for which we have no current evidence of testing".
667 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
668 if present. Previously it was ignored.
670 JH/24 Start using specified-initialisers in C structure init coding. This is
671 a C99 feature (it's 2017, so now considered safe).
673 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
674 if was a fixed-sized field and bitmask ops via macros; it is now more
677 PP/07 GitHub PR 56: Apply MariaDB build fix.
678 Patch provided by Jaroslav Škarvada.
680 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
681 during Coverity cleanups [4.87 JH/47]
682 Diagnosis and fix provided by Michael Fischer v. Mollard.
684 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
685 the right size to place the terminating semicolon on its own folded
686 line, the header hash was calculated to an incorrect value thanks to
687 the (relaxed) space the fold became.
689 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
690 and confused the parent.
692 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
693 which could crash as a result. This could lead to undeliverable messages.
695 JH/28 Logging: "next input sent too soon" now shows where input was truncated
698 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
699 matters on fast-turnover and PID-randomising systems, which were getting
700 out-of-order delivery.
702 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
703 a possibly-overlapping copy. The symptom was that "Remote host closed
704 connection in response to HELO" was logged instead of the actual 4xx
707 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
708 Previously only that bufferd was discarded, resulting in SYMTP command
711 JH/32 DKIM: when a message has multiple signatures matching an identity given
712 in dkim_verify_signers, run the dkim acl once for each. Previously only
713 one run was done. Bug 2189.
715 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
716 "panic the current process" to "deliberately defer". The panic log is
717 still written with the problem list name; the mail and reject logs now
718 get a temp-reject line for the message that was being handled, saying
719 something like "domains check lookup or other defer". The SMTP 451
720 message is still "Temporary local problem".
722 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
723 A crafted sequence of BDAT commands could result in in-use memory beeing
724 freed. CVE-2017-16943.
726 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
727 from SMTP input. Previously it was always done; now only done for DATA
728 and not BDAT commands. CVE-2017-16944.
730 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
731 to the message (such as an overlong header line). Previously this was
732 not done and we did not exit BDAT mode. Followon from the previous item
733 though a different problem.
739 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
740 than -2003 did; needs libidn2 in addition to libidn.
742 JH/02 The path option on a pipe transport is now expanded before use.
744 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
745 Patch provided by "Björn", documentation fix added too.
747 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
748 missing a wire-to-host endian conversion.
750 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
751 close after a BDAT command line could be taken as a following command,
752 giving a synch failure. Fix by only checking for synch immediately
753 before acknowledging the chunk.
755 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
756 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
757 macro. Patches provided by Josh Soref.
759 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
760 Previously we did not; the RFC seems ambiguous and VRFY is not listed
761 by IANA as a service extension. However, John Klensin suggests that we
764 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
765 the dkim code may be unix-mode line endings rather than smtp wire-format
766 CRLF, so prepend a CR to any bare LF.
768 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
769 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
771 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
772 the first were themselves being wrongly included in the feed into dkim
773 processing; with most chunk sizes in use this resulted in an incorrect
774 body hash calculated value.
776 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
777 DKIM signature block, for verification. Although advised against by
778 standards it is specifically not ruled illegal.
780 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
782 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
783 missing a body hash (the bh= tag).
785 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
786 It seems that HAProxy sends the Proxy Protocol information in clear and
787 only then does a TLS startup, so do the same.
789 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
790 TCP connections (such as for Spamd) unless the daemon successfully set
791 Fast Open mode on its listening sockets. This fixes breakage seen on
792 too-old kernels or those not configured for Fast Open, at the cost of
793 requiring both directions being enabled for TFO, and TFO never being used
794 by non-daemon-related Exim processes.
796 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
797 endings, at least on the first header line. Try to canonify any that get
798 past that check, despite the cost.
800 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
801 now limited to an arbitrary five deep, while parsing addresses with the
802 strip_excess_angle_brackets option enabled.
804 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
805 instead leave the unprompted TLS handshake in socket buffer for the
806 TLS library to consume.
808 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
810 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
812 JH/16 Drop variables when they go out of scope. Memory management drops a whole
813 region in one operation, for speed, and this leaves assigned pointers
814 dangling. Add checks run only under the testsuite which checks all
815 variables at a store-reset and panics on a dangling pointer; add code
816 explicitly nulling out all the variables discovered. Fixes one known
817 bug: a transport crash, where a dangling pointer for $sending_ip_address
818 originally assigned in a verify callout, is re-used.
820 PP/06 Drop '.' from @INC in various Perl scripts.
822 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
824 PP/08 Reduce a number of compilation warnings under clang; building with
825 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
826 should be warning-free.
828 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
830 HS/01 Fix portability problems introduced by PP/08 for platforms where
831 realloc(NULL) is not equivalent to malloc() [SunOS et al].
833 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
834 chunk. This allows us to accept broken chunked messages. We need a more
835 general solution here.
837 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
838 already-broken messages in the queue.
840 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
842 JH/19 Fix reference counting bug in routing-generated-address tracking.
848 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
849 supports it and a size is available (ie. the sending peer gave us one).
851 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
852 years of being deprecated). The replacements are the ACLs
853 acl_smtp_mime and acl_not_smtp_mime.
855 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
856 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
857 or even an in-clear connection were permitted. Now, if the host lookup
858 was dnssec and dane was requested then the host is only used if the
859 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
860 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
861 if one fails this test.
862 This means that a poorly-configured remote DNS will make it incommunicado;
863 but it protects against a DNS-interception attack on it.
865 JH/04 Bug 1810: make continued-use of an open smtp transport connection
866 non-noisy when a race steals the message being considered.
868 JH/05 If main configuration option tls_certificate is unset, generate a
869 self-signed certificate for inbound TLS connections.
871 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
872 in rewrites and routers.
874 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
875 and logged a warning sing 4.83; now they are a configuration file error.
877 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
878 (lacking @domain). Apply the same qualification processing as RCPT.
880 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
882 JH/10 Support ${sha256:} applied to a string (as well as the previous
885 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
886 a cutthrough deliver is pending, as we always want to make a connection.
887 This also avoids re-routing the message when later placing the cutthrough
888 connection after a verify cache hit.
889 Do not update it with the verify result either.
891 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
892 when routing results in more than one destination address.
894 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
895 signing (which inhibits the cutthrough capability). Previously only
896 the presence of an option was tested; now an expansion evaluating as
897 empty is permissible (obviously it should depend only on data available
898 when the cutthrough connection is made).
900 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
901 the relevant preceding SMTP command did not note the pipelining mode.
903 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
904 Previously they were not counted.
906 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
907 as one having no matching records. Previously we deferred the message
908 that needed the lookup.
910 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
911 distinguished as "(=".
913 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
914 for missing MX records. Previously it only worked for missing A records.
916 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
918 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
919 after the data-go-ahead and data-ack. Patch from Jason Betts.
921 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
922 even for a "none" policy. Patch from Tony Meyer.
924 JH/22 Fix continued use of a connection for further deliveries. If a port was
925 specified by a router, it must also match for the delivery to be
928 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
929 When one of the recipients of a message was unsuitable for the connection
930 (has no matching addresses), we lost track of needing to mark it
931 deferred. As a result mail would be lost.
933 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
935 JH/25 Decoding ACL controls is now done using a binary search; the source code
936 takes up less space and should be simpler to maintain. Merge the ACL
937 condition decode tables also, with similar effect.
939 JH/26 Fix problem with one_time used on a redirect router which returned the
940 parent address unchanged. A retry would see the parent address marked as
941 delivered, so not attempt the (identical) child. As a result mail would
944 JH/27 Fix a possible security hole, wherein a process operating with the Exim
945 UID can gain a root shell. Credit to http://www.halfdog.net/ for
946 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
949 JH/28 Enable {spool,log} filesystem space and inode checks as default.
950 Main config options check_{log,spool}_{inodes,space} are now
951 100 inodes, 10MB unless set otherwise in the configuration.
953 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
954 Previously it only applied to the main-section connection policy
957 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
959 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
960 by me. Added RFC7919 DH primes as an alternative.
962 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
963 are not in the system include path.
965 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
966 GnuTLS, when a session startup failed (eg because the client disconnected)
967 Exim did stdio operations after fclose. This was exposed by a recent
968 change which nulled out the file handle after the fclose.
970 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
971 signed directly by the cert-signing cert, rather than an intermediate
972 OCSP-signing cert. This is the model used by LetsEncrypt.
974 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
976 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
977 an incoming connection.
979 HS/02 Bug 1802: Do not half-close the connection after sending a request
982 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
983 fallback to "prime256v1".
985 JH/34 SECURITY: Use proper copy of DATA command in error message.
986 Could leak key material. Remotely exploitable. CVE-2016-9963.
992 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
993 and 3.4.4 - once the server is enabled to respond to an OCSP request
994 it does even when not requested, resulting in a stapling non-aware
995 client dropping the TLS connection.
997 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
998 support variable-length bit vectors. No functional change.
1000 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1001 The I= interface field on outgoing lines is now after the H= remote
1002 host field, same as incoming lines. There is a separate
1003 outgoing_interface log selector which allows you to disable the
1006 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1007 If not running log_selector +smtp_connection the mainlog would be held
1008 open indefinitely after a "too many connections" event, including to a
1009 deleted file after a log rotate. Leave the per net connection logging
1010 leaving it open for efficiency as that will be quickly detected by the
1011 check on the next write.
1013 HS/01 Bug 1671: Fix post transport crash.
1014 Processing the wait-<transport> messages could crash the delivery
1015 process if the message IDs didn't exist for some reason. When
1016 using 'split_spool_directory=yes' the construction of the spool
1017 file name failed already, exposing the same netto behaviour.
1019 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1020 mime_regex ACL conditions.
1022 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1023 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1024 to HELO, local diagnostic string.
1026 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1027 log line to debug. Even when configured with a tls authenticator many
1028 client connections are expected to not authenticate in this way, so
1029 an authenticate fail is not an error.
1031 HS/02 Add the Exim version string to the process info. This way exiwhat
1032 gives some more detail about the running daemon.
1034 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1035 matter for fast-change records such as DNSBLs.
1037 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1038 retry record, even if constant. There may be multiple transports with
1039 different interface settings and the retry behaviour needs to be kept
1042 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1044 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1046 JH/10 Bug 840: fix log_defer_output option of pipe transport
1048 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1049 hosts_randomize. This matters a lot when combined with mua_wrapper.
1051 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1052 ${quote_pgsql:<string>} operator.
1054 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1056 JH/14 Change header file name for freeradius-client. Relevant if compiling
1057 with Radius support; from the Gentoo tree and checked under Fedora.
1059 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1061 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1062 option transport_filter, meaning no filtering.
1064 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1066 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1067 defaults to "*" (all hosts). The variable is now available when not built
1068 with TLS, default unset, mainly to enable keeping the testsuite sane.
1069 If a server certificate is not supplied (via tls_certificate) an error is
1070 logged, and clients will find TLS connections fail on startup. Presumably
1071 they will retry in-clear.
1072 Packagers of Exim are strongly encouraged to create a server certificate
1073 at installation time.
1075 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1076 with the $config_file variable.
1078 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1079 in transport context, after the attempt, and per-recipient. The latter type
1080 is per host attempted. The event data is the error message, and the errno
1081 information encodes the lookup type (A vs. MX) used for the (first) host,
1082 and the trailing two digits of the smtp 4xx response.
1084 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1085 to write to mainlog (or rejectlog, paniclog) in the window between file
1086 creation and permissions/ownership being changed. Particularly affects
1087 installations where exicyclog is run as root, rather than exim user;
1088 result is that the running daemon panics and dies.
1090 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1092 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1093 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1094 "pri" and "weight". Note that the previous implicit priority given by the
1095 list order is no longer honoured.
1097 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1098 for DKIM processing.
1100 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1101 by defining SUPPORT_SOCKS.
1103 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1104 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1105 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1106 variables are renamed to proxy_{local,external}_{address,port}.
1108 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1109 for a build by defining SUPPORT_I18N
1111 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1112 of the query string, and make ${quote_redis:} do that quoting.
1114 JH/29 Move Events support from Experimental to mainline, enabled by default
1115 and removable for a build by defining DISABLE_EVENT.
1117 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1119 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1120 cached by the daemon.
1122 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1123 by defining LOOKUP_REDIS. The libhiredis library is required.
1125 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1126 keys are given for lookup.
1128 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1129 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1130 only supported when built with TLS support. The PolarSSL SHA routines
1131 are still used when the TLS library is too old for convenient support.
1133 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1134 openssl_options), for security. OpenSSL forces this from version 1.1.0
1135 server-side so match that on older versions.
1137 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1138 allocation for $value could be released as the expansion processing
1139 concluded, but leaving the global pointer active for it.
1141 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1142 and to use the domains and local_parts ACL conditions.
1144 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1145 incorrectly not doubled on cutthrough transmission, hence seen as a
1146 body-termination at the receiving system - resulting in truncated mails.
1147 Commonly the sender saw a TCP-level error, and retransmitted the message
1148 via the normal store-and-forward channel. This could result in duplicates
1149 received - but deduplicating mailstores were liable to retain only the
1150 initial truncated version.
1152 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1154 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1156 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1157 we're in there, support oversigning also; bug 1309.
1159 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1161 HS/04 Add support for keep_environment and add_environment options.
1163 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1164 either intentional arithmetic overflow during PRNG, or testing config-
1167 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1168 delivery resulted in actual delivery. Cancel cutthrough before DATA
1171 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1172 rejects a recipient: pass the reject to the originator.
1174 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1175 Many were false-positives and ignorable, but it's worth fixing the
1178 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1179 for the new environment-manipulation done at startup. Move the routines
1180 from being local to tls.c to being global via the os.c file.
1182 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1183 an extract embedded as result-arg for a map, the first arg for extract
1184 is unavailable so we cannot tell if this is a numbered or keyed
1185 extraction. Accept either.
1191 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1194 JH/02 The smtp transport option "multi_domain" is now expanded.
1196 JH/03 The smtp transport now requests PRDR by default, if the server offers
1199 JH/04 Certificate name checking on server certificates, when exim is a client,
1200 is now done by default. The transport option tls_verify_cert_hostnames
1201 can be used to disable this per-host. The build option
1202 EXPERIMENTAL_CERTNAMES is withdrawn.
1204 JH/05 The value of the tls_verify_certificates smtp transport and main options
1205 default to the word "system" to access the system default CA bundle.
1206 For GnuTLS, only version 3.0.20 or later.
1208 JH/06 Verification of the server certificate for a TLS connection is now tried
1209 (but not required) by default. The verification status is now logged by
1210 default, for both outbound TLS and client-certificate supplying inbound
1213 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1216 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1217 Status Notification (bounce) messages are now MIME format per RFC 3464.
1218 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1219 under the control of the dsn_advertise_hosts option, and routers may
1220 have a dsn_lasthop option.
1222 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1223 default, modifiable by a malware= option. The list separator for
1224 the options can now be changed in the usual way. Bug 68.
1226 JH/10 The smtp_receive_timeout main option is now expanded before use.
1228 JH/11 The incoming_interface log option now also enables logging of the
1229 local interface on delivery outgoing connections.
1231 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1232 if the interface and destination host and port all match.
1234 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1237 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1238 Patch from Andrew Lewis.
1240 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1241 now supports optional time-restrictions, weighting, and priority
1242 modifiers per server. Patch originally by <rommer@active.by>.
1244 JH/16 The spamd_address main option now supports a mixed list of local
1245 and remote servers. Remote servers can be IPv6 addresses, and
1246 specify a port-range.
1248 JH/17 Bug 68: The spamd_address main option now supports an optional
1249 timeout value per server.
1251 JH/18 Bug 1581: Router and transport options headers_add/remove can
1252 now have the list separator specified.
1254 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1257 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1260 JH/21 Support for the A6 type of dns record is withdrawn.
1262 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1263 rather than the verbs used.
1265 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1266 from 255 to 1024 chars.
1268 JH/24 Verification callouts now attempt to use TLS by default.
1270 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1271 are generic router options now. The defaults didn't change.
1273 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1274 Original patch from Alexander Shikoff, worked over by JH.
1276 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1277 files if ZCAT_COMMAND is not executable.
1279 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1281 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1283 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1284 Normally benign, it bites when the pair was led to by a CNAME;
1285 modern usage is to not canonicalize the domain to a CNAME target
1286 (and we were inconsistent anyway for A-only vs AAAA+A).
1288 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1290 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1291 when evaluating $sender_host_dnssec.
1293 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1294 $sender_helo_dnssec variable.
1296 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1298 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1300 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1302 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1303 documented as working, but never had. Support all but $spam_report.
1305 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1306 added for tls authenticator.
1308 HS/03 Add perl_taintmode main config option
1314 TL/01 When running the test suite, the README says that variables such as
1315 no_msglog_check are global and can be placed anywhere in a specific
1316 test's script, however it was observed that placement needed to be near
1317 the beginning for it to behave that way. Changed the runtest perl
1318 script to read through the entire script once to detect and set these
1319 variables, reset to the beginning of the script, and then run through
1320 the script parsing/test process like normal.
1322 TL/02 The BSD's have an arc4random API. One of the functions to induce
1323 adding randomness was arc4random_stir(), but it has been removed in
1324 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1325 function when detected.
1327 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1328 cause callback expansion.
1330 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1331 syntax errors in an expansion can be treated as a string instead of
1332 logging or causing an error, due to the internal use of bool_lax
1333 instead of bool when processing it.
1335 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1336 server certificates when making smtp deliveries.
1338 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1340 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1342 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1344 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1345 Merged patch from Sebastian Wiedenroth.
1347 JH/05 Fix results-pipe from transport process. Several recipients, combined
1348 with certificate use, exposed issues where response data items split
1349 over buffer boundaries were not parsed properly. This eventually
1350 resulted in duplicates being sent. This issue only became common enough
1351 to notice due to the introduction of connection certificate information,
1352 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1354 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1355 size buffer was used, resulting in syntax errors when an expansion
1358 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1359 version 3.3.6 or later.
1361 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1362 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1363 both become "event_action", the variables become $event_name, $event_data
1364 and $event_defer_errno. There is a new variable $verify_mode, usable in
1365 routers, transports and related events. The tls:cert event is now also
1366 raised for inbound connections, if the main configuration event_action
1369 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1370 early OCSP support, but no stapling (appears to be less than 1.0.0).
1372 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1373 server certificate names available under the smtp transport option
1374 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1377 JH/10 Time-related extraction expansions from certificates now use the main
1378 option "timezone" setting for output formatting, and are consistent
1379 between OpenSSL and GnuTLS compilations. Bug 1541.
1381 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1382 encoded parameter in the incoming message. Bug 1558.
1384 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1385 include certificate info, eximon was claiming there were spoolfile
1388 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1390 JH/14 Log delivery-related information more consistently, using the sequence
1391 "H=<name> [<ip>]" wherever possible.
1393 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1394 are problematic for Debian distribution, omit them from the release
1397 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1399 JH/16 Fix string representation of time values on 64bit time_t architectures.
1402 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1403 output list separator was used.
1408 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1409 checkers that were complaining about end of non-void function with no
1412 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1413 This was a regression introduced in 4.83 by another bugfix.
1415 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1417 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1418 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1424 TF/01 Correctly close the server side of TLS when forking for delivery.
1426 When a message was received over SMTP with TLS, Exim failed to clear up
1427 the incoming connection properly after forking off the child process to
1428 deliver the message. In some situations the subsequent outgoing
1429 delivery connection happened to have the same fd number as the incoming
1430 connection previously had. Exim would try to use TLS and fail, logging
1431 a "Bad file descriptor" error.
1433 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1434 utilities have not been installed.
1436 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1437 temporary space as the ACL may create new global variables.
1439 TL/01 LDAP support uses per connection or global context settings, depending
1440 upon the detected version of the libraries at build time.
1442 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1443 to extract and use the src ip:port in logging and expansions as if it
1444 were a direct connection from the outside internet. PPv2 support was
1445 updated based on HAProxy spec change in May 2014.
1447 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1449 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1450 Properly escape header and check for NULL return.
1452 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1455 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1457 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1458 characters in header names, implemented as a verify condition.
1459 Contributed by Michael Fischer v. Mollard.
1461 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1462 results permerror and temperror. Previous values are deprecated but
1463 still accepted. In a future release, err_perm and err_temp will be
1464 completely removed, which will be a backward incompatibility if the
1465 ACL tests for either of these two old results. Patch contributed by
1466 user bes-internal on the mailing list.
1468 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1470 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1471 selectors, in both main and reject logs.
1473 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1476 JH/07 Add malware type "sock" for talking to simple daemon.
1478 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1480 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1481 routers/transports under cutthrough routing.
1483 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1484 numbers. Touch up "bool" conditional to keep the same definition.
1486 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1488 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1490 JH/12 Expand items in router/transport headers_add or headers_remove lists
1491 individually rather than the list as a whole. Bug 1452.
1493 Required for reasonable handling of multiple headers_ options when
1494 they may be empty; requires that headers_remove items with embedded
1495 colons must have them doubled (or the list-separator changed).
1497 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1498 view the policy declared in the DMARC record. Currently, $dmarc_status
1499 is a combined value of both the record presence and the result of the
1502 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1504 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1505 dnslookup router and the smtp transport (applying to the forward
1508 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1509 of ldap servers used for a specific lookup. Patch provided by Heiko
1512 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1513 New variable $lookup_dnssec_authenticated for observability.
1515 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1516 Patch submitted by Lars Timman.
1518 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1520 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1521 Requires trusted mode and valid format message id, aborts otherwise.
1522 Patch contributed by Heiko Schlichting.
1524 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1525 certextract with support for various fields. Bug 1358.
1527 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1528 is requested by default, modifiable by smtp transport option
1531 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1532 operate on certificate variables to give certificate fingerprints
1533 Also new ${sha256:cert_variable}.
1535 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1537 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1538 Christian Aistleitner.
1540 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1542 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1543 file. Patch from Wolfgang Breyha.
1545 JH/25 Expand the coverage of the delivery $host and $host_address to
1546 client authenticators run in verify callout. Bug 1476.
1548 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1549 align with daemon_smtp_ports. Bug 72.
1551 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1552 support and error reporting did not work properly.
1554 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1555 and is readable. Patch from Andrew Colin Kissa.
1557 TL/14 Enhance documentation of ${run expansion and how it parses the
1558 commandline after expansion, particularly in the case when an
1559 unquoted variable expansion results in an empty value.
1561 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1563 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1564 by renaming away from it. Addresses GH issue 15, reported by
1567 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1569 TL/15 SECURITY: prevent double expansion in math comparison functions
1570 (can expand unsanitized data). Not remotely exploitable.
1577 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1579 PP/02 Make -n do something, by making it not do something.
1580 When combined with -bP, the name of an option is not output.
1582 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1585 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1586 $sender_host_name and config options to manage this, and basic check
1589 PP/05 DSCP support for outbound connections and control modifier for inbound.
1591 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1592 (Only plugin which currently uses this is kerberos4, which nobody should
1593 be using, but we should make it available and other future plugins might
1594 conceivably use it, even though it would break NAT; stuff *should* be
1595 using channel bindings instead).
1597 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1598 name; added for Sendmail compatibility; requires admin caller.
1599 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1600 just ignore it); requires trusted caller.
1601 Also parse but ignore: -Ac -Am -X<logfile>
1604 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1606 TL/02 Add +smtp_confirmation as a default logging option.
1608 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1609 Patch by Magnus Holmgren from 2007-02-20.
1611 TL/04 Bugzilla 1281 - Spec typo.
1612 Bugzilla 1283 - Spec typo.
1613 Bugzilla 1290 - Spec grammar fixes.
1615 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1617 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1619 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1620 mailing list by by Dmitry Isaikin.
1622 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1624 JH/02 Support "G" suffix to numbers in ${if comparisons.
1626 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1628 NM/01 Bugzilla 1197 - Spec typo
1629 Bugzilla 1196 - Spec examples corrections
1631 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1633 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1634 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1637 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1638 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1640 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1641 "acl {{name}{arg}...}", and optional args on acl condition
1644 JH/05 Permit multiple router/transport headers_add/remove lines.
1646 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1648 JH/07 Avoid using a waiting database for a single-message-only transport.
1649 Performance patch from Paul Fisher. Bugzilla 1262.
1651 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1654 JH/09 Add $headers_added variable, with content from use of ACL modifier
1655 add_header (but not yet added to the message). Bugzilla 199.
1657 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1658 Pulled from Bugzilla 817 by Wolfgang Breyha.
1660 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1662 (nb: this is the same fix as in Exim 4.80.1)
1664 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1667 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1670 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1672 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1673 advertise SMTP AUTH mechanism to us, instead of a generic
1674 protocol violation error. Also, make Exim more robust to bad
1675 data from the Dovecot auth socket.
1677 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1679 When a queue runner is handling a message, Exim first routes the
1680 recipient addresses, during which it prunes them based on the retry
1681 hints database. After that it attempts to deliver the message to
1682 any remaining recipients. It then updates the hints database using
1685 So if a recipient address works intermittently, it can get repeatedly
1686 deferred at routing time. The retry hints record remains fresh so the
1687 address never reaches the final cutoff time.
1689 This is a fairly common occurrence when a user is bumping up against
1690 their storage quota. Exim had some logic in its local delivery code
1691 to deal with this. However it did not apply to per-recipient defers
1692 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1694 This change adds a proper retry rule check during routing so that the
1695 final cutoff time is checked against the message's age. We only do
1696 this check if there is an address retry record and there is not a
1697 domain retry record; this implies that previous attempts to handle
1698 the address had the retry_use_local_parts option turned on. We use
1699 this as an approximation for the destination being like a local
1700 delivery, as in LMTP.
1702 I suspect this new check makes the old local delivery cutoff check
1703 redundant, but I have not verified this so I left the code in place.
1705 TF/02 Correct gecos expansion when From: is a prefix of the username.
1707 Test 0254 submits a message to Exim with the header
1711 When I ran the test suite under the user fanf2, Exim expanded
1712 the header to contain my full name, whereas it should have added
1713 a Resent-Sender: header. It erroneously treats any prefix of the
1714 username as equal to the username.
1716 This change corrects that bug.
1718 GF/01 DCC debug and logging tidyup
1719 Error conditions log to paniclog rather than rejectlog.
1720 Debug lines prefixed by "DCC: " to remove any ambiguity.
1722 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1724 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1725 Bug spotted by Jeremy Harris; was flawed since initial commit.
1726 Would have resulted in OCSP responses post-SNI triggering an Exim
1727 NULL dereference and crash.
1729 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1731 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1732 Bug detection, analysis and fix by Samuel Thibault.
1733 Bugzilla 1331, Debian bug #698092.
1735 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1737 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1738 Server implementation by Todd Lyons, client by JH.
1739 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1740 config variable "prdr_enable" controls whether the server
1741 advertises the facility. If the client requests PRDR a new
1742 acl_data_smtp_prdr ACL is called once for each recipient, after
1743 the body content is received and before the acl_smtp_data ACL.
1744 The client is controlled by both of: a hosts_try_prdr option
1745 on the smtp transport, and the server advertisement.
1746 Default client logging of deliveries and rejections involving
1747 PRDR are flagged with the string "PRDR".
1749 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1750 fclose(). Diagnosis by Todd Lyons.
1752 PP/17 Update configure.default to handle IPv6 localhost better.
1753 Patch by Alain Williams (plus minor tweaks).
1756 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1757 This is now consistent with GnuTLS, and is now documented: the
1758 previous undocumented portable approach to treating the option as
1759 unset was to force an expansion failure. That still works, and
1760 an empty string is now equivalent.
1762 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1763 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1764 not performing validation itself.
1766 PP/20 Added force_command boolean option to pipe transport.
1767 Patch from Nick Koston, of cPanel Inc.
1769 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1772 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1774 PP/21 Fix eximon continuous updating with timestamped log-files.
1775 Broken in a format-string cleanup in 4.80, missed when I repaired the
1776 other false fix of the same issue.
1777 Report and fix from Heiko Schlichting.
1780 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1781 Report from Prashanth Katuri.
1783 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1784 It's SecureTransport, so affects any MacOS clients which use the
1785 system-integrated TLS libraries, including email clients.
1787 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1788 using a MIME ACL for non-SMTP local injection.
1789 Report and assistance in diagnosis by Warren Baker.
1791 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1793 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1795 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1796 last id that failed so it may be referenced in subsequent ACL's.
1798 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1801 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1802 ldap library initialization, allowing self-signed CA's to be
1803 used. Also properly sets require_cert option later in code by
1804 using NULL (global ldap config) instead of ldap handle (per
1805 session). Bug diagnosis and testing by alxgomz.
1807 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1808 the src/util/ subdirectory.
1810 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1811 renamed to Transport Post Delivery Action by Jeremy Harris, as
1814 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1815 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1816 redis_servers = needs to be configured which will be used by the redis
1817 lookup. Patch from Warren Baker, of The Packet Hub.
1819 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1821 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1822 hostname or reverse DNS when processing a host list. Used suggestions
1823 from multiple comments on this bug.
1825 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1827 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1828 Missed a few lines, added it to make the runtest require no keyboard
1831 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1832 contains upper case chars. Make router use caseful_local_part.
1834 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1835 support when GnuTLS has been built with p11-kit.
1841 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1843 This, or similar/improved, will also be change PP/11 of 4.82.
1849 PP/01 Handle short writes when writing local log-files.
1850 In practice, only affects FreeBSD (8 onwards).
1851 Bugzilla 1053, with thanks to Dmitry Isaikin.
1853 NM/01 Bugzilla 949 - Documentation tweak
1855 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1858 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1860 PP/02 Implemented gsasl authenticator.
1862 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1864 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1865 `pkg-config foo` for cflags/libs.
1867 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1868 with rest of GSASL and with heimdal_gssapi.
1870 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1871 `pkg-config foo` for cflags/libs for the TLS implementation.
1873 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1874 properties get this fed in as external SSF. A number of robustness
1875 and debugging improvements to the cyrus_sasl authenticator.
1877 PP/08 cyrus_sasl server now expands the server_realm option.
1879 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1880 Patch by Jeremy Harris.
1882 PP/10 Added dbmjz lookup type.
1884 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1886 PP/12 MAIL args handles TAB as well as SP, for better interop with
1887 non-compliant senders.
1888 Analysis and variant patch by Todd Lyons.
1890 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1891 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1892 Patch from Dirk Mueller <dmueller@suse.com>
1894 PP/13 tls_peerdn now print-escaped for spool files.
1895 Observed some $tls_peerdn in wild which contained \n, which resulted
1896 in spool file corruption.
1898 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1899 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1900 or write after TLS renegotiation, which otherwise led to messages
1903 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1904 as a tracking header (ie: a signed header comes before the signature).
1905 Patch from Wolfgang Breyha.
1907 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1908 comma-sep list; embedded commas doubled.
1910 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1912 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1914 Report and patch from Dmitry Banschikov.
1916 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1917 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1918 failures. We appear to now support TLS1.1+ with Exim.
1920 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1921 lets Exim select keys and certificates based upon TLS SNI from client.
1922 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1923 before an outbound SMTP session. New log_selector, +tls_sni.
1925 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1926 NULL dereference. Report and patch from Alun Jones.
1928 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1929 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1930 Not seeing resolver debug output on NetBSD, but suspect this is a
1931 resolver implementation change.
1933 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1934 Left warnings. Added "eximon gdb" invocation mode.
1936 PP/21 Defaulting "accept_8bitmime" to true, not false.
1938 PP/22 Added -bw for inetd wait mode support.
1940 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1941 locate the relevant includes and libraries. Made this the default.
1943 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1944 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1946 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1947 This may cause build issues on older platforms.
1949 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1950 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1951 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1952 Added SNI support via GnuTLS too.
1953 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1955 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1957 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1958 Applied second patch from Janne, implementing suggestion to default
1959 multiple-strings-in-record handling to match SPF spec.
1961 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1963 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1964 read-only, out of scope).
1965 Patch from Wolfgang Breyha, report from Stuart Northfield.
1967 PP/29 Fix three issues highlighted by clang analyser static analysis.
1968 Only crash-plausible issue would require the Cambridge-specific
1969 iplookup router and a misconfiguration.
1970 Report from Marcin Mirosław.
1972 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1974 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1975 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1976 As part of this, removing so much warning spew let me fix some minor
1977 real issues in debug logging.
1979 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1980 assignment on my part. Fixed.
1982 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1983 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1984 Janne Snabb (who went above and beyond: thank you).
1986 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1987 string otherwise requires a connection and a bunch more work and it's
1988 relatively easy to get wrong. Should also expose TLS library linkage
1991 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1992 64-bit ${eval} (JH/03).
1994 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1995 GNU libc to support some of the 64-bit stuff, should not lead to
1996 conflicts. Defined before os.h is pulled in, so if a given platform
1997 needs to override this, it can.
1999 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2000 protection layer was required, which is not implemented.
2001 Bugzilla 1254, patch from Wolfgang Breyha.
2003 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2004 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2005 tls_dhparam take prime identifiers. Also unbreak combination of
2006 OpenSSL+DH_params+TLSSNI.
2008 PP/39 Disable SSLv2 by default in OpenSSL support.
2014 PP/01 Solaris build fix for Oracle's LDAP libraries.
2015 Bugzilla 1109, patch from Stephen Usher.
2017 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2019 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2022 TF/02 Fix a couple more cases where we did not log the error message
2023 when unlink() failed. See also change 4.74-TF/03.
2025 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2026 lock up or crash if it happened to be inside a call to libc when it
2027 got a SIGUSR1 from exiwhat.
2029 The SIGUSR1 handler appends the current process status to the process
2030 log which is later printed by exiwhat. It used to use the general
2031 purpose logging code to do this, but several functions it calls are
2032 not safe for signals.
2034 The new output code in the SIGUSR1 handler is specific to the process
2035 log, and simple enough that it's easy to inspect for signal safety.
2036 Removing some special cases also simplifies the general logging code.
2037 Removing the spurious timestamps from the process log simplifies
2040 TF/04 Improved ratelimit ACL condition.
2042 The /noupdate option has been deprecated in favour of /readonly which
2043 has clearer semantics. The /leaky, /strict, and /readonly update modes
2044 are mutually exclusive. The update mode is no longer included in the
2045 database key; it just determines when the database is updated. (This
2046 means that when you upgrade Exim will forget old rate measurements.)
2048 Exim now checks that the per_* options are used with an update mode that
2049 makes sense for the current ACL. For example, when Exim is processing a
2050 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2051 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2052 must specify per_mail/readonly. If you omit the update mode it defaults to
2053 /leaky where that makes sense (as before) or /readonly where required.
2055 The /noupdate option is now undocumented but still supported for
2056 backwards compatibility. It is equivalent to /readonly except that in
2057 ACLs where /readonly is required you may specify /leaky/noupdate or
2058 /strict/noupdate which are treated the same as /readonly.
2060 A useful new feature is the /count= option. This is a generalization
2061 of the per_byte option, so that you can measure the throughput of other
2062 aggregate values. For example, the per_byte option is now equivalent
2063 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2065 The per_rcpt option has been generalized using the /count= mechanism
2066 (though it's more complicated than the per_byte equivalence). When it is
2067 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2068 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2069 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2070 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2071 Note that using per_rcpt with a non-readonly update mode in more than
2072 one ACL will cause the recipients to be double-counted. (The per_mail
2073 and per_byte options don't have this problem.)
2075 The handling of very low rates has changed slightly. If the computed rate
2076 is less than the event's count (usually one) then this event is the first
2077 after a long gap. In this case the rate is set to the same as this event's
2078 count, so that the first message of a spam run is counted properly.
2080 The major new feature is a mechanism for counting the rate of unique
2081 events. The new per_addr option counts the number of different
2082 recipients that someone has sent messages to in the last time period. It
2083 behaves like per_rcpt if all the recipient addresses are different, but
2084 duplicate recipient addresses do not increase the measured rate. Like
2085 the /count= option this is a general mechanism, so the per_addr option
2086 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2087 example, measure the rate that a client uses different sender addresses
2088 with the options per_mail/unique=$sender_address. There are further
2089 details in the main documentation.
2091 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2093 TF/06 Removed a few PCRE remnants.
2095 TF/07 Automatically extract Exim's version number from tags in the git
2096 repository when doing development or release builds.
2098 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2099 Bugzilla 879. Patch from Paul Fisher.
2101 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2102 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2105 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2107 PP/05 Variable $av_failed, true if the AV scanner deferred.
2108 Bugzilla 1078. Patch from John Horne.
2110 PP/06 Stop make process more reliably on build failure.
2111 Bugzilla 1087. Patch from Heiko Schlittermann.
2113 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2114 Bugzilla 1089. Patch from Heiko Schlittermann.
2116 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2117 Bugzilla 1131. Patch from Holger Weiß.
2119 PP/09 Handle IPv6 addresses with SPF.
2120 Bugzilla 860. Patch from Wolfgang Breyha.
2122 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2124 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2127 PP/11 match_* no longer expand right-hand-side by default.
2128 New compile-time build option, EXPAND_LISTMATCH_RHS.
2129 New expansion conditions, "inlist", "inlisti".
2131 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2133 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2135 PP/14 fix log_write() format string regression from TF/03.
2136 Bugzilla 1152. Patch from Dmitry Isaikin.
2142 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2144 PP/02 Harmonised TLS library version reporting; only show if debugging.
2145 Layout now matches that introduced for other libraries in 4.74 PP/03.
2147 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2149 PP/04 New "dns_use_edns0" global option.
2151 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2154 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2155 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2157 TK/01 Updated PolarSSL code to 0.14.2.
2158 Bugzilla 1097. Patch from Andreas Metzler.
2160 PP/07 Catch divide-by-zero in ${eval:...}.
2161 Fixes bugzilla 1102.
2163 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2166 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2167 format-string attack -- SECURITY: remote arbitrary code execution.
2169 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2170 time unintentionally subject to list matching rules, letting the header
2171 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2172 arbitrary string expansion). This allowed for information disclosure.
2174 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2175 INT_MIN/-1 -- value coerced to INT_MAX.
2181 NM/01 Workaround for PCRE version dependency in version reporting
2184 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2185 This fixes portability to compilers other than gcc, notably
2186 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2188 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2189 makefiles for portability to HP-UX and POSIX correctness.
2191 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2192 Also via indented variable definition in the Makefile.
2193 (Debugging by Oliver Heesakkers).
2195 PP/02 Restore caching of spamd results with expanded spamd_address.
2196 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2198 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2199 Improves build reliability. Fix from: Frank Elsner
2201 NM/02 Fix wide character breakage in the rfc2047 coding
2202 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2204 NM/03 Allow underscore in dnslist lookups
2205 Fixes bug 1026. Patch from Graeme Fowler
2207 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2208 Code patches from Adam Ciarcinski of NetBSD.
2210 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2213 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2214 is logged, to avoid truncation. Patch from John Horne.
2216 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2217 Patch from Jakob Hirsch.
2219 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2220 SQL string expansion failure details.
2221 Patch from Andrey Oktyabrski.
2223 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2224 Patch from Simon Arlott.
2226 PP/09 New lookups functionality failed to compile on old gcc which rejects
2227 extern declarations in function scope.
2228 Patch from Oliver Fleischmann
2230 PP/10 Use sig_atomic_t for flags set from signal handlers.
2231 Check getgroups() return and improve debugging.
2232 Fixed developed for diagnosis in bug 927 (which turned out to be
2235 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2236 Patch from Mark Zealey.
2238 PP/12 Bugzilla 1056: Improved spamd server selection.
2239 Patch from Mark Zealey.
2241 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2242 Based on patch from Heiko Schlittermann.
2244 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2245 Patch from Uwe Doering, sign-off by Michael Haardt.
2247 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2248 variable declaration deep within a block. Bug and patch from
2251 PP/15 lookups-Makefile IRIX compatibility coercion.
2253 PP/16 Make DISABLE_DKIM build knob functional.
2255 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2256 Patch by Simon Arlott
2258 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2259 variable argument macros. Our copy now differs from upstream.
2265 TF/01 Failure to get a lock on a hints database can have serious
2266 consequences so log it to the panic log.
2268 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2269 controlled using the smtp_confirmation log selector.
2271 TF/03 Include the error message when we fail to unlink a spool file.
2273 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2274 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2275 for maintaining out-of-tree patches for some time.
2277 PP/01 Bugzilla 139: Documentation and portability issues.
2278 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2279 Handle per-OS dynamic-module compilation flags.
2281 PP/02 Let /dev/null have normal permissions.
2282 The 4.73 fixes were a little too stringent and complained about the
2283 permissions on /dev/null. Exempt it from some checks.
2284 Reported by Andreas M. Kirchwitz.
2286 PP/03 Report version information for many libraries, including
2287 Exim version information for dynamically loaded libraries. Created
2288 version.h, now support a version extension string for distributors
2289 who patch heavily. Dynamic module ABI change.
2291 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2292 privilege escalation vulnerability whereby the Exim run-time user
2293 can cause root to append content of the attacker's choosing to
2296 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2299 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2300 If dropping privileges for untrusted macros, we disabled normal logging
2301 on the basis that it would fail; for the Exim run-time user, this is not
2302 the case, and it resulted in successful deliveries going unlogged.
2303 Fixed. Reported by Andreas Metzler.
2309 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2310 only prepend for the Resent-* case. Fixes regression introduced in
2311 Exim 4.70 by NM/22 for Bugzilla 607.
2313 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2314 increasing numbers of administrators be bitten by this.
2316 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2318 PP/03 Bugzilla 994: added openssl_options main configuration option.
2320 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2322 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2324 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2326 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2327 without a peer certificate, leading to a segfault because of an
2328 assumption that peers always have certificates. Be a little more
2329 paranoid. Problem reported by Martin Tscholak.
2331 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2332 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2333 NB: ClamAV planning to remove STREAM in "middle of 2010".
2334 CL also introduces -bmalware, various -d+acl logging additions and
2335 more caution in buffer sizes.
2337 PP/09 Implemented reverse_ip expansion operator.
2339 PP/10 Bugzilla 937: provide a "debug" ACL control.
2341 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2343 PP/12 Bugzilla 973: Implement --version.
2345 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2347 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2349 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2351 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2352 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2353 ignore trailing whitespace.
2355 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2357 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2360 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2361 Notification from Dr Andrew Aitchison.
2363 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2364 ExtendedDetectionInfo response format.
2365 Notification from John Horne.
2367 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2370 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2371 XSL and documented dependency on system catalogs, with examples of how
2374 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2377 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2378 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2379 configuration file which is writeable by the Exim user or group.
2381 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2382 of configuration files to cover files specified with the -C option if
2383 they are going to be used with root privileges, not just the default
2386 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2387 option (effectively making it always true).
2389 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2390 files to be used while preserving root privileges.
2392 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2393 that rogue child processes cannot use them.
2395 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2396 run-time user, instead of root.
2398 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2399 Exim run-time user without dropping privileges.
2401 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2402 result string, instead of calling string_vformat() twice with the same
2405 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2406 for other users. Others should always drop root privileges if they use
2407 -C on the command line, even for a whitelisted configure file.
2409 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2411 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2417 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2418 $data_path, and $header_path variables; fixed documentation bugs and
2421 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2422 exipick to access non-standard spools, including the "frozen" queue
2425 NM/01 Bugzilla 965: Support mysql stored procedures.
2426 Patch from Alain Williams
2428 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2430 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2431 Patch from Andreas Metzler
2433 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2434 Patch from Kirill Miazine
2436 NM/05 Bugzilla 671: Added umask to procmail example.
2438 JJ/03 installed exipick 20100323.0, fixing doc bug
2440 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2441 directory. Notification and patch from Dan Rosenberg.
2443 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2445 TK/02 Improve log output when DKIM signing operation fails.
2447 MH/01 Treat the transport option dkim_domain as a colon separated
2448 list, not as a single string, and sign the message with each element,
2449 omitting multiple occurences of the same signer.
2451 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2452 Bugzilla 985, 986. Patch by Simon Arlott
2454 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2455 Patch by Simon Arlott
2457 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2458 MBX locking. Notification from Dan Rosenberg.
2464 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2466 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2468 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2470 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2472 NM/04 Bugzilla 915: Flag broken perl installation during build.
2478 TK/01 Added patch by Johannes Berg that expands the main option
2479 "spamd_address" if it starts with a dollar sign.
2481 TK/02 Write list of recipients to X-Envelope-Sender header when building
2482 the mbox-format spool file for content scanning (suggested by Jakob
2485 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2486 (http://www.dcc-servers.net/) support via dccifd. Activated by
2487 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2489 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2490 by Mark Daniel Reidel <mr@df.eu>.
2492 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2493 When building exim an external PCRE library is now needed -
2494 PCRE is a system library on the majority of modern systems.
2495 See entry on PCRE_LIBS in EDITME file.
2497 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2498 conversation. Added nologin parameter to request.
2499 Patch contributed by Kirill Miazine.
2501 TF/01 Do not log submission mode rewrites if they do not change the address.
2503 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2505 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2506 log files in place. Contributed by Roberto Lima.
2508 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2510 TF/03 Bugzilla 615: When checking the local_parts router precondition
2511 after a local_part_suffix or local_part_prefix option, Exim now
2512 does not use the address's named list lookup cache, since this
2513 contains cached lookups for the whole local part.
2515 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2516 Robert Millan. Documentation is in experimental-spec.txt.
2518 TF/04 Bugzilla 668: Fix parallel build (make -j).
2520 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2522 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2523 Patch provided by Jan Srzednicki.
2525 TF/05 Leading white space used to be stripped from $spam_report which
2526 wrecked the formatting. Now it is preserved.
2528 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2529 that they are available at delivery time.
2531 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2533 TF/08 TLS error reporting now respects the incoming_interface and
2534 incoming_port log selectors.
2536 TF/09 Produce a more useful error message if an SMTP transport's hosts
2537 setting expands to an empty string.
2539 NM/06 Bugzilla 744: EXPN did not work under TLS.
2540 Patch provided by Phil Pennock.
2542 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2543 Patch provided by Richard Godbee.
2545 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2546 acl_smtp_notquit, added index entry.
2548 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2549 Patch provided by Eugene Bujak.
2551 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2552 accept(). Patch provided by Maxim Dounin.
2554 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2555 Patch provided by Phil Pennock.
2557 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2559 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2560 Patch provided by Brad "anomie" Jorsch.
2562 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2563 Patch provided by Dean Brooks.
2565 TK/05 Add native DKIM support (does not depend on external libraries).
2567 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2568 Patch provided by Graeme Fowler.
2570 NM/16 Bugzilla 851: Documentation example syntax fix.
2572 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2574 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2577 NM/19 Bugzilla 745: TLS version reporting.
2578 Patch provided by Phil Pennock.
2580 NM/20 Bugzilla 167: bool: condition support.
2581 Patch provided by Phil Pennock.
2583 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2584 clients. Patch provided by Phil Pennock.
2586 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2587 Patch provided by Brad "anomie" Jorsch.
2589 NM/23 Bugzilla 687: Fix misparses in eximstats.
2590 Patch provided by Heiko Schlittermann.
2592 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2593 Patch provided by Heiko Schlittermann.
2595 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2596 plus update to original patch.
2598 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2600 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2601 Patch provided by David Brownlee.
2603 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2605 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2607 NM/30 Bugzilla 888: TLS documentation bugfixes.
2609 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2611 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2612 Unlike the original bugzilla I have changed all shell scripts in src tree.
2614 NM/33 Bugzilla 898: Transport filter timeout fix.
2615 Patch by Todd Rinaldo.
2617 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2618 Patch by Serge Demonchaux.
2620 NM/35 Bugzilla 39: Base64 decode bug fixes.
2621 Patch by Jakob Hirsch.
2623 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2625 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2627 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2629 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2635 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2636 ALT-N's libdkim that I have put here:
2637 http://duncanthrax.net/exim-experimental/
2639 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2640 were called 'true' and it seems that C99 defines that as a reserved
2641 keyword to be used with 'bool' variable types. That means you could
2642 not include C99-style headers which use bools without triggering
2643 build errors in sieve.c.
2645 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2646 as mailq or other aliases. Changed the --help handling significantly
2647 to do whats expected. exim_usage() emits usage/help information.
2649 SC/01 Added the -bylocaldomain option to eximstats.
2651 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2653 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2655 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2661 PH/01 Another patch from the Sieve maintainer.
2663 PH/02 When an IPv6 address is converted to a string for single-key lookup
2664 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2665 dots are used instead of colons so that keys in lsearch files need not
2666 contain colons. This was done some time before quoting was made available
2667 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2668 (notated using the quote facility) so as to distinguish them from IPv4
2669 keys. This meant that lookups for IP addresses in host lists did not work
2670 for iplsearch lookups.
2672 This has been fixed by arranging for IPv6 addresses to be expressed with
2673 colons if the lookup type is iplsearch. This is not incompatible, because
2674 previously such lookups could never work.
2676 The situation is now rather anomalous, since one *can* have colons in
2677 ordinary lsearch keys. However, making the change in all cases is
2678 incompatible and would probably break a number of configurations.
2680 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2683 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2684 conversion specification without a maximum field width, thereby enabling
2685 a rogue spamd server to cause a buffer overflow. While nobody in their
2686 right mind would setup Exim to query an untrusted spamd server, an
2687 attacker that gains access to a server running spamd could potentially
2688 exploit this vulnerability to run arbitrary code as the Exim user.
2690 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2691 $primary_hostname instead of what libspf2 thinks the hosts name is.
2693 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2694 a directory entry by the name of the lookup key. Previously, if a
2695 symlink pointed to a non-existing file or a file in a directory that
2696 Exim lacked permissions to read, a lookup for a key matching that
2697 symlink would fail. Now it is enough that a matching directory entry
2698 exists, symlink or not. (Bugzilla 503.)
2700 PH/03 The body_linecount and body_zerocount variables are now exported in the
2703 PH/04 Added the $dnslist_matched variable.
2705 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2706 This means they are set thereafter only if the connection becomes
2709 PH/06 Added the client_condition to authenticators so that some can be skipped
2710 by clients under certain conditions.
2712 PH/07 The error message for a badly-placed control=no_multiline_responses left
2713 "_responses" off the end of the name.
2715 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2717 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2718 (without spaces) instead of just copying the configuration text.
2720 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2722 PH/11 Added $max_received_linelength.
2724 PH/12 Added +ignore_defer and +include_defer to host lists.
2726 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2727 way in which PCRE > 7.0 is built.
2729 PH/14 Implemented queue_only_load_latch.
2731 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2732 MAIL command. The effect was to mangle the value on 64-bit systems.
2734 PH/16 Another patch from the Sieve maintainer.
2736 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2738 PH/18 If a system quota error occurred while trying to create the file for
2739 a maildir delivery, the message "Mailbox is full" was not appended to the
2740 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2741 only to a quota excession during the actual writing of the file.
2743 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2744 characters?) which causes problems in log lines. The DN values are now
2745 passed through string_printing() before being added to log lines.
2747 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2748 and InterBase are left for another time.)
2750 PH/21 Added message_body_newlines option.
2752 PH/22 Guard against possible overflow in moan_check_errorcopy().
2754 PH/23 POSIX allows open() to be a macro; guard against that.
2756 PH/24 If the recipient of an error message contained an @ in the local part
2757 (suitably quoted, of course), incorrect values were put in $domain and
2758 $local_part during the evaluation of errors_copy.
2764 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2765 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2768 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2769 issue a MAIL command.
2771 PH/02 In an ACL statement such as
2773 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2775 if a client was not listed at all, or was listed with a value other than
2776 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2777 the condition was not true (as it should be), so access was not denied.
2778 The bug was that the ! inversion was incorrectly passed on to the second
2779 item. This has been fixed.
2781 PH/03 Added additional dnslists conditions == and =& which are different from
2782 = and & when the dns lookup returns more than one IP address.
2784 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2785 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2787 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2788 FSYNC, which compiles an option called disable_fsync that allows for
2789 bypassing fsync(). The documentation is heavily laced with warnings.
2791 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2793 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2794 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2795 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2796 including adding "make clean"; (3) Added -fPIC when compiling the test
2797 dynamically loaded module, to get rid of a warning.
2799 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2800 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2801 = 0s. The bug is otherwise harmless.
2803 PH/07 There was a bug in the dovecot authenticator such that the value of
2804 $auth1 could be overwritten, and so not correctly preserved, after a
2805 successful authentication. This usually meant that the value preserved by
2806 the server_setid option was incorrect.
2808 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2810 PH/09 Installed PCRE release 7.0.
2812 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2813 run for batched SMTP input. It is now run at the start of every message
2814 in the batch. While fixing this I discovered that the process information
2815 (output by running exiwhat) was not always getting set for -bs and -bS
2816 input. This is fixed, and it now also says "batched" for BSMTP.
2818 PH/11 Added control=no_pipelining.
2820 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2821 patch, slightly modified), and move the expansion of helo_data till after
2822 the connection is made in the smtp transport (so it can use these
2825 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2827 PH/14 Added log_selector = +pid.
2829 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2831 PH/16 Add ${if forany and ${if forall.
2833 PH/17 Added dsn_from option to vary the From: line in DSNs.
2835 PH/18 Flush SMTP output before performing a callout, unless control =
2836 no_callout_flush is set.
2838 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2839 was true (the default) a successful delivery failed to delete the retry
2840 item, thus causing premature timeout of the address. The bug is now
2843 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2845 PH/21 Long custom messages for fakedefer and fakereject are now split up
2846 into multiline responses in the same way that messages for "deny" and
2847 other ACL rejections are.
2849 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2850 with slight modification.
2852 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2853 draft, changing the syntax and factoring some duplicate code".
2855 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2856 for deliveries of the second and subsequent messages over the same SMTP
2859 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2860 ${reduce, with only minor "tidies".
2862 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2864 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2865 expansion side effects.
2867 PH/27 When a message times out after an over-quota error from an Exim-imposed
2868 quota, the bounce message says "mailbox is full". This message was not
2869 being given when it was a system quota that was exceeded. It now should
2872 MH/03 Made $recipients available in local_scan(). local_scan() already has
2873 better access to the recipient list through recipients_list[], but
2874 $recipients can be useful in postmaster-provided expansion strings.
2876 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2877 in the case of a MAIL command with additional options following the
2878 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2879 were accidentally chopped off.
2881 PH/29 SMTP synchronization checks are implemented when a command is read -
2882 there is a check that no more input is waiting when there shouldn't be
2883 any. However, for some commands, a delay in an ACL can mean that it is
2884 some time before the response is written. In this time, more input might
2885 arrive, invalidly. So now there are extra checks after an ACL has run for
2886 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2887 pipelining has not been advertised.
2889 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2891 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2892 returned in the SMTP response when smtp_return_error_details was false.
2893 This has been fixed.
2895 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2896 instead of the C I/O that was originally supplied, because problems were
2897 reported on Solaris.
2899 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2900 Exim which did not show up earlier: it was assuming that a call to
2901 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2902 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2903 was a macro that became an assignment, so it seemed to work. This has
2904 changed to a proper function call with a void return, hence the compile
2905 error. Exim's code has been fixed.
2907 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2910 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2912 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2914 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2915 because it turns out to be too big or there is a timeout) there is no
2916 "Completed" line in the log. When some messages of this type were
2917 selected by exigrep, they were listed as "not completed". Others were
2918 picked up by some special patterns. I have improved the selection
2919 criteria to be more general.
2921 PH/38 The host_find_failed option in the manualroute router can now be set
2922 to "ignore", to completely ignore a host whose IP address cannot be
2923 found. If all hosts are ignored, the behaviour is controlled by the new
2924 host_all_ignored option.
2926 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2927 homing or because of multiple MX records with /mx) generated more than
2928 one IP address, and the following item turned out to be the local host,
2929 all the secondary addresses of the first item were incorrectly removed
2930 from the list, along with the local host and any following hosts (which
2931 is what is supposed to happen).
2933 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2934 whoever called Exim into the -H file. In the case of the daemon it was
2935 behaving confusingly. When first started, it used values for whoever
2936 started the daemon, but after a SIGHUP it used the Exim user (because it
2937 calls itself on a restart). I have changed the code so that it now always
2940 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2941 message are rejected with the same error (e.g. no authentication or bad
2942 sender address), and a DATA command is nevertheless sent (as can happen
2943 with PIPELINING or a stupid MUA), the error message that was given to the
2944 RCPT commands is included in the rejection of the DATA command. This is
2945 intended to be helpful for MUAs that show only the final error to their
2948 PH/42 Another patch from the Sieve maintainer.
2950 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2951 Eximstats - Fixed some broken HTML links and added missing column headers
2953 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2954 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2956 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2958 PH/43 Yet another patch from the Sieve maintainer.
2960 PH/44 I found a way to check for a TCP/IP connection going away before sending
2961 the response to the final '.' that terminates a message, but only in the
2962 case where the client has not sent further data following the '.'
2963 (unfortunately, this is allowed). However, in many cases there won't be
2964 any further data because there won't be any more messages to send. A call
2965 to select() can be used: if it shows that the input is "ready", there is
2966 either input waiting, or the socket has been closed. An attempt to read
2967 the next input character can distinguish the two cases. Previously, Exim
2968 would have sent an OK response which the client would never have see.
2969 This could lead to message repetition. This fix should cure that, at
2970 least in a lot of common cases.
2972 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2973 advertised in response to EHLO.
2979 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2980 fixed by 4.65/MH/01 (is this a record?) are fixed:
2982 (i) An empty string was always treated as zero by the numeric comparison
2983 operators. This behaviour has been restored.
2985 (ii) It is documented that the numeric comparison operators always treat
2986 their arguments as decimal numbers. This was broken in that numbers
2987 starting with 0 were being interpreted as octal.
2989 While fixing these problems I realized that there was another issue that
2990 hadn't been noticed. Values of message_size_limit (both the global option
2991 and the transport option) were treated as octal if they started with 0.
2992 The documentation was vague. These values are now always treated as
2993 decimal, and I will make that clear in the documentation.
2999 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3000 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3003 MH/01 Don't check that the operands of numeric comparison operators are
3004 integers when their expansion is in "skipping" mode (fixes bug
3005 introduced by 4.64-PH/07).
3007 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3008 child addresses, Exim now panics and dies. Previously, because the count
3009 is held in a short int, deliveries were likely to be lost. As such a
3010 large number of recipients for a single message is ridiculous
3011 (performance will be very, very poor), I have chosen to impose a limit
3012 rather than extend the field.
3018 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3019 leftover -K file (the existence of which was triggered by #402).
3020 While we were at it, introduced process PID as part of the -K
3021 filename. This should rule out race conditions when creating
3024 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3025 processing considerably. Previous code took too long for large mails,
3026 triggering a timeout which in turn triggers #401.
3028 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3029 in the DK code in transports.c. sendfile() is not really portable,
3030 hence the _LINUX specificness.
3032 TF/01 In the add_headers option to the mail command in an Exim filter,
3033 there was a bug that Exim would claim a syntax error in any
3034 header after the first one which had an odd number of characters
3037 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3038 callout verification, Exim cached a "reject" for the entire domain. This
3039 is correct for most verifications, but it is not correct for a recipient
3040 verification with use_sender or use_postmaster set, because in that case
3041 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3042 case of MAIL FROM:<> rejection from other early rejections (e.g.
3043 rejection of HELO). When verifying a recipient using a non-null MAIL
3044 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3045 Whatever the result of the callout, the value of the domain cache is
3046 left unchanged (for any other kind of callout, getting as far as trying
3047 RCPT means that the domain itself is ok).
3049 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3052 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3053 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3054 session, the daemon ignores ECONNECT errors and logs others; it now
3055 ignores EPIPE as well.
3057 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3058 (quoted-printable decoding).
3060 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3061 later the small subsequent patch to fix an introduced bug.
3063 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3065 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3067 PH/08 An error is now given if message_size_limit is specified negative.
3069 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3070 to be given (somewhat) arbitrary names.
3072 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3075 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3076 miscellaneous code fixes
3078 PH/10 Added the log_reject_target ACL modifier to specify where to log
3081 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3082 hostname. This is wrong, because it relates to the incoming message (and
3083 probably the interface on which it is arriving) and not to the outgoing
3084 callout (which could be using a different interface). This has been
3085 changed to use the value of the helo_data option from the smtp transport
3086 instead - this is what is used when a message is actually being sent. If
3087 there is no remote transport (possible with a router that sets up host
3088 addresses), $smtp_active_hostname is used.
3090 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3091 tweaks were necessary in order to get it to work (see also 21 below):
3092 (a) The code assumed that strncpy() returns a negative number on buffer
3093 overflow, which isn't the case. Replaced with Exim's string_format()
3095 (b) There were several signed/unsigned issues. I just did the minimum
3096 hacking in of casts. There is scope for a larger refactoring.
3097 (c) The code used strcasecmp() which is not a standard C function.
3098 Replaced with Exim's strcmpic() function.
3099 (d) The code set only $1; it now sets $auth1 as well.
3100 (e) A simple test gave the error "authentication client didn't specify
3101 service in request". It would seem that Dovecot has changed its
3102 interface. Fortunately there's a specification; I followed it and
3103 changed what the client sends and it appears to be working now.
3105 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3108 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3109 address A is aliased to B and C, where B exists and C does not. Without
3110 -v the output is "A verified" because verification stops after a
3111 successful redirection if more than one address is generated. However,
3112 with -v the child addresses are also verified. Exim was outputting "A
3113 failed to verify" and then showing the successful verification for C,
3114 with its parentage. It now outputs "B failed to verify", showing B's
3115 parentage before showing the successful verification of C.
3117 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3118 look up a TXT record in a specific list after matching in a combined
3121 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3122 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3123 they consult the DNS. I had assumed they would set it the way they
3124 wanted; and indeed my experiments on Linux seem to show that in some
3125 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3126 To be on the safe side, however, I have now made the interface to
3127 host_find_byname() similar to host_find_bydns(), with an argument
3128 containing the DNS resolver options. The host_find_byname() function now
3129 sets these options at its start, just as host_find_bydns() does. The smtp
3130 transport options dns_qualify_single and dns_search_parents are passed to
3131 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3132 of host_find_byname() use the default settings of RES_DEFNAMES
3133 (qualify_single) but not RES_DNSRCH (search_parents).
3135 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3136 spool_read_header() do less string testing, by means of a preliminary
3137 switch on the second character of optional "-foo" lines. (This is
3138 overdue, caused by the large number of possibilities that now exist.
3139 Originally there were few.) While I was there, I also converted the
3140 str(n)cmp tests so they don't re-test the leading "-" and the first
3141 character, in the hope this might squeeze out yet more improvement.
3143 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3144 flag allowing group syntax was set by the header_syntax check but not
3145 turned off, possible causing trouble later; (2) The flag was not being
3146 set at all for the header_verify test, causing "group"-style headers to
3147 be rejected. I have now set it in this case, and also caused header_
3148 verify to ignore an empty address taken from a group. While doing this, I
3149 came across some other cases where the code for allowing group syntax
3150 while scanning a header line wasn't quite right (mostly, not resetting
3151 the flag correctly in the right place). These bugs could have caused
3152 trouble for malformed header lines. I hope it is now all correct.
3154 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3155 with the "reply" argument non-NULL. The code, however (which originally
3156 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3157 but it didn't always do it. This confused somebody who was copying the
3158 code for some other use. I have removed all the tests.
3160 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3161 feature that was used to support insecure browsers during the U.S. crypto
3162 embargo. It requires special client support, and Exim is probably the
3163 only MTA that supported it -- and would never use it because real RSA is
3164 always available. This code has been removed, because it had the bad
3165 effect of slowing Exim down by computing (never used) parameters for the
3166 RSA_EXPORT functionality.
3168 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3169 authenticator to fail if there's a tab character in the incoming data
3170 (there should never be unless someone is messing about, as it's supposed
3171 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3172 if the connection is using TLS or if the remote IP is the same as the
3173 local IP, and the "valid-client-cert option" if a client certificate has
3176 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3177 authenticators. This can be used for authorization after authentication
3178 succeeds. (In the case of plaintext, it servers for both authentication
3181 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3182 if any retry times were supplied.
3184 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3185 connection, where there is no client IP address to check. In this
3186 situation, the verify now always succeeds.
3188 PH/25 Applied John Jetmore's -Mset patch.
3190 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3192 PH/27 In a string expansion for a processed (not raw) header when multiple
3193 headers of the same name were present, leading whitespace was being
3194 removed from all of them, but trailing whitespace was being removed only
3195 from the last one. Now trailing whitespace is removed from each header
3196 before concatenation. Completely empty headers in a concatenation (as
3197 before) are ignored.
3199 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3200 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3202 PH/29 [Removed. This was a change that I later backed out, and forgot to
3203 correct the ChangeLog entry (that I had efficiently created) before
3204 committing the later change.]
3206 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3207 address errors (4xx response to RCPT) over the same connection as other
3208 messages routed to the same hosts. Such deliveries are always "forced",
3209 so retry times are not inspected. This resulted in far too many retries
3210 for the affected addresses. The effect occurred only when there were more
3211 hosts than the hosts_max_try setting in the smtp transport when it had
3212 the 4xx errors. Those hosts that it had tried were not added to the list
3213 of hosts for which the message was waiting, so if all were tried, there
3214 was no problem. Two fixes have been applied:
3216 (i) If there are any address or message errors in an SMTP delivery, none
3217 of the hosts (tried or untried) are now added to the list of hosts
3218 for which the message is waiting, so the message should not be a
3219 candidate for sending over the same connection that was used for a
3220 successful delivery of some other message. This seems entirely
3221 reasonable: after all the message is NOT "waiting for some host".
3222 This is so "obvious" that I'm not sure why it wasn't done
3223 previously. Hope I haven't missed anything, but it can't do any
3224 harm, as the worst effect is to miss an optimization.
3226 (ii) If, despite (i), such a delivery is accidentally attempted, the
3227 routing retry time is respected, so at least it doesn't keep
3228 hammering the server.
3230 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3231 in ${readsocket because some servers need this prod.
3233 PH/32 Added some extra debug output when updating a wait-xxx database.
3235 PH/33 The hint "could be header name not terminated by colon", which has been
3236 given for certain expansion errors for a long time, was not being given
3237 for the ${if def:h_colon_omitted{... case.
3239 PH/34 The spec says: "With one important exception, whenever a domain list is
3240 being scanned, $domain contains the subject domain." There was at least
3241 one case where this was not true.
3243 PH/35 The error "getsockname() failed: connection reset by peer" was being
3244 written to the panic log as well as the main log, but it isn't really
3245 panic-worthy as it just means the connection died rather early on. I have
3246 removed the panic log writing for the ECONNRESET error when getsockname()
3249 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3250 runs only) independently of the message's sender address. This meant
3251 that, if the 4xx error was in fact related to the sender, a different
3252 message to the same recipient with a different sender could confuse
3253 things. In particular, this can happen when sending to a greylisting
3254 server, but other circumstances could also provoke similar problems.
3255 I have changed the default so that the retry time for these errors is now
3256 based a combination of the sender and recipient addresses. This change
3257 can be overridden by setting address_retry_include_sender=false in the
3260 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3261 remote server are returned as part of bounce messages. This was not
3262 happening for LMTP over a pipe (the lmtp transport), but now it is the
3263 same for both kinds of LMTP.
3265 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3266 in header lines that were in fact CNAMEs. This is no longer the case.
3268 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3269 and queue runs started by the daemon processed all messages. This has
3270 been fixed so that -R and -S can now usefully be given with -q<time>.
3272 PH/40 Import PCRE release 6.7 (fixes some bugs).
3274 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3276 PH/42 Give an error if -q is specified more than once.
3278 PH/43 Renamed the variables $interface_address and $interface_port as
3279 $received_ip_address and $received_port, to make it clear that these
3280 values apply to message reception, and not to the outgoing interface when
3281 a message is delivered. (The old names remain recognized, of course.)
3283 PH/44 There was no timeout on the connect() call when using a Unix domain
3284 socket in the ${readsocket expansion. There now is.
3286 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3287 be meaningful with "accept".
3289 SC/01 Eximstats V1.43
3290 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3292 SC/02 Eximstats V1.44
3293 Use a glob alias rather than an array ref in the generated
3294 parser. This improves both readability and performance.
3296 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3297 Collect SpamAssassin and rejection statistics.
3298 Don't display local sender or destination tables unless
3299 there is data to show.
3300 Added average volumes into the top table text output.
3302 SC/04 Eximstats V1.46
3303 Collect data on the number of addresses (recipients)
3304 as well as the number of messages.
3306 SC/05 Eximstats V1.47
3307 Added 'Message too big' to the list of mail rejection
3308 reasons (thanks to Marco Gaiarin).
3310 SC/06 Eximstats V1.48
3311 Mainlog lines which have GMT offsets and are too short to
3312 have a flag are now skipped.
3314 SC/07 Eximstats V1.49 (Alain Williams)
3315 Added the -emptyok flag.
3317 SC/08 Eximstats V1.50
3318 Fixes for obtaining the IP address from reject messages.
3320 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3321 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3322 whitespace changes from 4.64-PH/27
3324 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3327 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3328 are found, allow negative numbers in numeric criteria)
3330 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3332 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3333 to match changes made in 4.64-PH/43
3335 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3337 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3338 "domain" is actually an IP address, and force a failure. This locks out
3339 those revolvers/nameservers that support "A-for-A" lookups, in
3340 contravention of the specifications.
3342 PH/48 When a host name was looked up from an IP address, and the subsequent
3343 forward lookup of the name timed out, the host name was left in
3344 $sender_host_name, contrary to the specification.
3346 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3347 restricted to single-key lookups, Exim was not diagnosing an error if
3348 * or *@ was used with a query-style lookup.
3350 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3352 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3353 long ago, but noone interested enough thought of it. Let's just say that
3354 the "1.1" means that there are some new functions that weren't there at
3355 some point in the past.
3357 PH/51 Error processing for expansion failure of helo_data from an smtp
3358 transport during callout processing was broken.
3360 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3361 tested/used via the -bh/-bhc/-bs options.
3363 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3364 bug, fixed in subsequent PCRE releases).
3366 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3367 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3369 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3375 SC/01 Use a glob alias rather than an array ref in eximstats generated
3376 parser. This improves both readability and performance.
3378 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3379 Don't display local sender or destination tables in eximstats unless
3380 there is data to show.
3381 Added average volumes into the eximstats top table text output.
3383 SC/03 Collect data on the number of addresses (recipients) as well
3384 as the number of messages in eximstats.
3386 TF/01 Correct an error in the documentation for the redirect router. Exim
3387 does (usually) call initgroups() when daemonizing.
3389 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3390 with consistent privilege compared to when running as a daemon.
3392 TF/03 Note in the spec that $authenticated_id is not set for local
3393 submissions from trusted users.
3395 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3396 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3398 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3399 by adding some example configuration directives to the default
3400 configuration file. A little bit of work is required to uncomment the
3401 directives and define how usernames and passwords are checked, but
3402 there is now a framework to start from.
3404 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3405 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3406 without this. I don't know how relevant this is to other LDAP libraries.
3408 PH/02 Add the verb name to the "unknown ACL verb" error.
3410 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3412 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3414 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3415 directory not expanded when it should be if an expanded home directory
3416 was set for the address (which is overridden by the transport).
3418 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3421 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3422 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3423 because it is too late at that time, and has no effect.
3425 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3426 security issue with \' (bugzilla #107). I could not use the
3427 PQescapeStringConn() function, because it needs a PGconn value as one of
3430 PH/08 When testing addresses using -bt, indicate those final addresses that
3431 are duplicates that would not cause an additional delivery. At least one
3432 person was confused, thinking that -bt output corresponded to deliveries.
3433 (Suppressing duplicates isn't a good idea as you lose the information
3434 about possibly different redirections that led to the duplicates.)
3436 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3437 systems where poll() doesn't work, in particular OS X.
3439 PH/10 Added more information to debugging output for retry time not reached.
3441 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3442 operations in malware.c.
3444 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3447 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3448 syslog_duplication set false, log lines that would normally be written
3449 both the the main log and to the reject log were not written to syslog at
3452 PH/14 In the default configuration, change the use of "message" in ACL warn
3453 statements to "add_header".
3455 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3456 not followed by a command (e.g. "seen endif").
3458 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3459 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3462 PH/17 Added extra conditions to the default value of delay_warning_condition
3466 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3467 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3468 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3471 The Auto-Submitted: and various List- headers are standardised, whereas I
3472 don't think Precedence: ever was.
3474 PH/18 Refactored debugging code in route_finduser() to show more information,
3475 in particular, the error code if getpwnam() issues one.
3477 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3478 This is apparently needed in addition to the PH/07 change above to avoid
3479 any possible encoding problems.
3481 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3482 but not after initializing Perl.
3484 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3485 output them only if debugging. By default they are written stderr,
3486 apparently, which is not desirable.
3488 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3491 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3494 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3496 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3497 authenticated or an ident call has been made. Suppress the default
3498 values for $authenticated_id and $authenticated_sender (but permit -oMai
3499 and -oMas) when testing with -bh.
3501 PH/24 Re-jigged the order of the tests in the default configuration so that the
3502 tests for valid domains and recipients precede the DNS black list and CSA
3503 tests, on the grounds that those ones are more expensive.
3505 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3506 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3507 This bug exists in every version of Exim that I still have, right back to
3510 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3511 However, an attempt to turn on case-sensitivity in a regex key by
3512 including (?-i) didn't work because the subject string was already
3513 lowercased, and the effects were non-intuitive. It turns out that a
3514 one-line patch can be used to allow (?-i) to work as expected.
3520 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3521 other effects) broke the use of negated acl sub-conditions.
3523 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3526 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3527 "Deny" causes Exim to reject the incoming connection with a 554 error.
3528 Unfortunately, if there is a major crisis, such as a disk failure,
3529 tcp-wrappers gives "deny", whereas what one would like would be some
3530 kind of temporary error. A kludge has been added to help with this.
3531 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3532 554 error is used if errno is still zero or contains ENOENT (which occurs
3533 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3536 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3538 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3539 errors. Otherwise a message that provokes a temporary error (when other
3540 messages do not) can cause a whole host to time out.
3542 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3543 addresses were routed directly to files or pipes from a redirect router.
3544 File deliveries just didn't batch; pipe deliveries might have suffered
3547 PH/06 A failure to get a lock for a hints database would erroneously always say
3548 "Failed to get write lock", even when it was really a read lock.
3550 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3551 of 0600. This has been changed to use the value of the lockfile_mode
3552 option (which defaults to 0600).
3554 PH/08 Applied small patch from the Sieve maintainer.
3556 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3557 folder from quota calculations, a direct delivery into this folder messed
3558 up the contents of the maildirsize file. This was because the regex was
3559 used only to exclude .Trash (or whatever) when the size of the mailbox
3560 was calculated. There was no check that a delivery was happening into an
3561 excluded directory. This bug has been fixed by ignoring all quota
3562 processing for deliveries into excluded directories.
3564 PH/10 Added the maildirfolder_create_regex option to appendfile.
3570 PH/01 The code for finding all the local interface addresses on a FreeBSD
3571 system running IPv6 was broken. This may well have applied to all BSD
3572 systems, as well as to others that have similar system calls. The broken
3573 code found IPv4 interfaces correctly, but gave incorrect values for the
3574 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3575 that it would not match correctly against @[] and not recognize the IPv6
3578 PH/02 The ipliteral router was not recognizing addresses of the form user@
3579 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3581 PH/03 Added disable_ipv6.
3583 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3584 decoded form, because it is most often used to construct To: headers
3585 lines in autoreplies, and the decoded form may well be syntactically
3586 invalid. However, $reply_address has leading white space removed, and all
3587 newlines turned into spaces so that the autoreply transport does not
3590 PH/05 If group was specified without a user on a router, and no group or user
3591 was specified on a transport, the group from the router was ignored.
3593 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3594 for visible compile-time settings that can be used to change these
3595 numbers, for those that want even more. Backwards compatibility with old
3596 spool files has been maintained. However, going back to a previous Exim
3597 release will lost any variables that are in spool files.
3599 PH/07 Two small changes when running in the test harness: increase delay when
3600 passing a TCP/IP connection to a new process, in case the original
3601 process has to generate a bounce, and remove special handling of
3602 127.0.0.2 (sic), which is no longer necessary.
3604 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3605 be the same on different OS.
3607 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3610 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3611 whether --show-vars was specified or not
3613 JJ/02 exipick: Added support for new ACL variable spool format introduced
3616 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3617 syntactically invalid From: or Reply-to: line, and a filter used this to
3618 generate an autoreply, and therefore failed to obtain an address for the
3619 autoreply, Exim could try to deliver to a non-existent relative file
3620 name, causing unrelated and misleading errors. What now happens is that
3621 it logs this as a hard delivery error, but does not attempt to create a
3624 PH/11 The exinext utility has a -C option for testing purposes, but although
3625 the given file was scanned by exinext itself; it wasn't being passed on
3626 when Exim was called.
3628 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3629 an end-of-file indication when reading a command response.
3631 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3632 compiled. In many other places in Exim, IPv6 addresses are always
3633 recognized, so I have changed this. It also means that IPv4 domain
3634 literals of the form [IPV4:n.n.n.n] are now always recognized.
3636 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3637 used if the router is not running as root, for example, when verifying at
3638 ACL time, or when using -bh. The debugging output from this situation was
3639 non-existent - all you got was a failure to exec. I have made two
3642 (a) Failures to set uid/gid, the current directory, or a process leader
3643 in a subprocess such as that created by queryprogram now generate
3644 suitable debugging output when -d is set.
3646 (b) The queryprogram router detects when it is not running as root,
3647 outputs suitable debugging information if -d is set, and then runs
3648 the subprocess without attempting to change uid/gid.
3650 PH/15 Minor change to Makefile for building test_host (undocumented testing
3653 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3654 additional section of a DNS packet that returns MX or SRV records.
3655 Instead, it always explicitly searches for A/AAAA records. This avoids
3656 major problems that occur when a DNS server includes only records of one
3657 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3658 fixed another bug: if SRV records were looked up and the corresponding
3659 address records were *not* found in the additional section, the port
3660 values from the SRV records were lost.
3662 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3663 using the correct key (the original address) when searching the retry
3664 rules in order to find which one to use for generating the retry hint.
3666 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3667 adding the default one. Similarly, if it contains a Reply-To: header, the
3668 errors_reply_to option, if set, is not used.
3670 PH/19 When calculating a retry time, Exim used to measure the "time since
3671 failure" by looking at the "first failed" field in the retry record. Now
3672 it does not use this if it is later than than the arrival time of the
3673 message. Instead it uses the arrival time. This makes for better
3674 behaviour in cases where some deliveries succeed, thus re-setting the
3675 "first failed" field. An example is a quota failure for a huge message
3676 when small messages continue to be delivered. Without this change, the
3677 "time since failure" will always be short, possible causing more frequent
3678 delivery attempts for the huge message than are intended.
3679 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3681 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3682 $1, $2, $3) because the numerical variables can be reset during some
3683 expansion items (e.g. "match"), thereby losing the authentication data.
3685 PH/21 Make -bV show the size of off_t variables so that the test suite can
3686 decide whether to run tests for quotas > 2G.
3688 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3689 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3690 filecount value is greater than 2G or if a quota value is greater than 2G
3691 on a system where the size of off_t is not greater than 4, a panic error
3694 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3695 never match. The debug and -bh output now contains an explicit error
3696 message indicating a malformed IPv4 address or mask.
3698 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3699 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3700 PH/23 above applies.
3702 PH/25 Do not write to syslog when running in the test harness. The only
3703 occasion when this arises is a failure to open the main or panic logs
3704 (for which there is an explicit test).
3706 PH/26 Added the /no_tell option to "control=freeze".
3708 PH/27 If a host name lookup failed very early in a connection, for example, if
3709 the IP address matched host_lookup and the reverse lookup yielded a name
3710 that did not have a forward lookup, an error message of the form "no IP
3711 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3712 could be logged. Now it outputs the IP address instead of "NULL".
3714 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3715 allows the sender and the authenticated sender to be set when
3716 submitting a message from within Exim. Since child_open_exim() is
3717 documented for local_scan(), the new function should be too.
3719 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3720 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3721 results in an empty string is now treated as unset.
3723 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3725 PH/31 Added sender_verify_fail logging option.
3727 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3728 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3729 tidied the source and removed it altogether.
3731 PH/33 When a queue run was abandoned because the load average was too high, a
3732 log line was always written; now it is written only if the queue_run log
3733 selector is set. In addition, the log line for abandonment now contains
3734 information about the queue run such as the pid. This is always present
3735 in "start" and "stop" lines but was omitted from the "abandon" line.
3737 PH/34 Omit spaces between a header name and the colon in the error message that
3738 is given when verify = headers_syntax fails (if there are lots of them,
3739 the message gets confusing).
3741 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3742 names, as there are now some PTR records that contain slashes. This check
3743 is only to protect against broken name servers that fall over on strange
3744 characters, so the fact that it applies to all lookups doesn't matter.
3746 PH/36 Now that the new test suite is complete, we can remove some of the
3747 special code in Exim that was needed for the old test suite. For example,
3748 sorting DNS records because real resolvers return them in an arbitrary
3749 order. The new test suite's fake resolver always returns records in the
3752 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3753 bounces) except when queue_only is set, to avoid logging races between
3754 the different processes.
3756 PH/38 Panic-die if .include specifies a non-absolute path.
3758 PH/39 A tweak to the "H" retry rule from its user.
3760 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3761 a label. They prevented compilation on older perls.
3763 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3764 a warning to be raised on newish perls.
3766 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3767 on queue. Changes to match documented behaviour of showing count of
3768 messages matching specified criteria.
3770 PH/40 Changed the default ident timeout from 30s to 5s.
3772 PH/41 Added support for the use of login_cap features, on those BSD systems
3773 that have them, for controlling the resources used by pipe deliveries.
3775 PH/42 The content-scanning code uses fopen() to create files in which to put
3776 message data. Previously it was not paying any attention to the mode of
3777 the files. Exim runs with umask(0) because the rest of the code creates
3778 files with open(), and sets the required mode explicitly. Thus, these
3779 files were ending up world-writeable. This was not a big issue, because,
3780 being within the spool directory, they were not world-accessible. I have
3781 created a function called modefopen, which takes an additional mode
3782 argument. It sets umask(777), creates the file, chmods it to the required
3783 mode, then resets the umask. All the relevant calls to fopen() in the
3784 content scanning code have been changed to use this function.
3786 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3787 to 24 hours. This avoids potential overflow problems when processing G
3788 and H retry rules. I suspect nobody ever tinkers with this value.
3790 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3792 PH/45 When the plaintext authenticator is running as a client, the server's
3793 challenges are checked to ensure they are valid base64 strings. By
3794 default, the authentication attempt is cancelled if an invalid string is
3795 received. Setting client_ignore_invalid_base64 true ignores these errors.
3796 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3797 they are received. Thus, the responses can be made to depend on the
3798 challenges. If an invalid string is ignored, an empty string is placed in
3801 PH/46 Messages that are created by the autoreply transport now contains a
3802 References: header, in accordance with RFCs 2822 and 3834.
3804 PH/47 Added authenticated_sender_force to the smtp transport.
3806 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3808 PH/49 Installed latest patch from the Sieve maintainer.
3810 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3811 was also set, the appendfile transport was unnecessarily scanning a
3812 directory of message files (e.g. for maildir delivery) to find the count
3813 of files (along with the size), even though it did not need this
3814 information. It now does the scan only if it needs to find either the
3815 size of the count of files.
3817 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3819 PH/52 Two bugs concerned with error handling when the smtp transport is
3822 (i) Exim was not creating retry information for temporary errors given
3823 for individual recipients after the DATA command when the smtp transport
3824 was used in LMTP mode. This meant that they could be retried too
3825 frequently, and not timed out correctly.
3827 (ii) Exim was setting the flag that allows error details to be returned
3828 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3829 recipients that were returned after the DATA command.
3831 PH/53 This is related to PH/52, but is more general: for any failing address,
3832 when detailed error information was permitted to be returned to the
3833 sender, but the error was temporary, then after the final timeout, only
3834 "retry timeout exceeded" was returned. Now it returns the full error as
3835 well as "retry timeout exceeded".
3837 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3838 do this, and (what is worse) MTAs that accept it.
3840 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3841 will now be deprecated.
3843 PH/56 New os.c-cygwin from the Cygwin maintainer.
3845 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3846 formats (previously only available in exim formats via -bpr, -bpru,
3847 and -bpra. Now also available in native and exiqgrep formats)
3849 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3850 with very large, slow to parse queues
3852 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3854 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3856 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3857 responses to authentication challenges, though it was showing the
3858 challenges; (ii) I've removed the CR characters from the debug output for
3861 PH/58 Allow for the insertion of a newline as well as a space when a string
3862 is turned into more than one encoded-word during RFC 2047 encoding. The
3863 Sieve code now uses this.
3865 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3866 data_4xx, lost_connection, tls_required.
3868 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3869 message was being sent as an SMTP response.
3871 PH/61 Add -l and -k options to exicyclog.
3873 PH/62 When verifying, if an address was redirected to one new address, so that
3874 verification continued, and the new address failed or deferred after
3875 having set something in $address_data, the value of $address_data was not
3876 passed back to the ACL. This was different to the case when no
3877 redirection occurred. The value is now passed back in both cases.
3879 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3880 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3881 use login_cap.h, so on its own it isn't the distinguishing feature. The
3882 new name refers directly to the setclassresources() function.
3884 PH/65 Added configuration files for NetBSD3.
3886 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3888 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3889 is preferred over IPv4.
3891 PH/68 The bounce_return_message and bounce_return_body options were not being
3892 honoured for bounces generated during the reception of non-SMTP messages.
3893 In particular, this applied to messages rejected by the ACL. This bug has
3894 been fixed. However, if bounce_return_message is true and bounce_return_
3895 body is false, the headers that are returned for a non-SMTP message
3896 include only those that have been read before the error was detected.
3897 (In the case of an ACL rejection, they have all been read.)
3899 PH/69 The HTML version of the specification is now built in a directory called
3900 spec_html instead of spec.html, because the latter looks like a path with
3901 a MIME-type, and this confuses some software.
3903 PH/70 Catch two compiler warnings in sieve.c.
3905 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3906 function verify_get_ident() calls ip_connect() to connect a socket, but
3907 if the "connect()" function timed out, ip_connect() used to close the
3908 socket. However, verify_get_ident() also closes the socket later, and in
3909 between Exim writes to the log, which may get opened at this point. When
3910 the socket was closed in ip_connect(), the log could get the same file
3911 descriptor number as the socket. This naturally causes chaos. The fix is
3912 not to close the socket in ip_connect(); the socket should be closed by
3913 the function that creates it. There was only one place in the code where
3914 this was missing, in the iplookup router, which I don't think anybody now
3915 uses, but I've fixed it anyway.
3917 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3918 well as to direct DNS lookups. Otherwise the handling of names in host
3919 lists is inconsistent and therefore confusing.
3925 PH/01 Two changes to the default runtime configuration:
3927 (1) Move the checks for relay_from_hosts and authenticated clients from
3928 after to before the (commented out) DNS black list checks.
3930 (2) Add control=submission to the relay_from_hosts and authenticated
3931 clients checks, on the grounds that messages accepted by these
3932 statements are most likely to be submissions.
3934 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3936 (1) Generate an error if the third argument for the ${prvs expansion is
3939 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3942 (3) Reset the variables that are obtained from the first argument of
3943 ${prvscheck and used in the second argument before leaving the code,
3944 because their memory is reclaimed, so using them afterwards may do
3947 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3948 one (it's much easier than Tom thought :-).
3950 (5) Because of (4), we can now allow for the use of $prvscheck_result
3951 inside the third argument.
3953 PH/03 For some reason, the default setting of PATH when running a command from
3954 a pipe transport was just "/usr/bin". I have changed it to
3957 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3958 anything to be listed in the output from -bV.
3960 PH/05 When a filter generated an autoreply, the entire To: header line was
3961 quoted in the delivery log line, like this:
3963 => >A.N.Other <ano@some.domain> <original@ddress> ...
3965 This has been changed so that it extracts the operative address. There
3966 may be more than one such address. If so, they are comma-separated, like
3969 => >ano@some.domain,ona@other.domain <original@ddress> ...
3971 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3972 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3973 not being looked up in the rDNS to get a host name, Exim was showing the
3974 IP address twice in Received: lines, even though the IP addresses were
3975 identical. For example:
3977 Received: from [1.2.3.4] (helo=[1.2.3.4])
3979 However, if the real host name was known, it was omitting the HELO data
3980 if it matched the actual IP address. This has been tidied up so that it
3981 doesn't show the same IP address twice.
3983 PH/07 When both +timestamp and +memory debugging was on, the value given by
3984 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3985 called by the debug printing, thereby overwriting the timestamp buffer.
3986 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3988 PH/08 When the original message was included in an autoreply transport, it
3989 always said "this is a copy of the message, including all the headers",
3990 even if body_only or headers_only was set. It now gives an appropriate
3993 PH/09 Applied a patch from the Sieve maintainer which:
3995 o fixes some comments
3996 o adds the (disabled) notify extension core
3997 o adds some debug output for the result of if/elsif tests
3998 o points to the current vacation draft in the documentation
3999 and documents the missing references header update
4003 o fixes a bug in processing the envelope test (when testing
4004 multiple envelope elements, the last element determined the
4007 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4008 Electronic Mail") by including:
4010 Auto-submitted: auto-generated
4012 in the messages that it generates (bounce messages and others, such as
4013 warnings). In the case of bounce messages for non-SMTP messages, there was
4014 also a typo: it was using "Auto_submitted" (underscore instead of
4015 hyphen). Since every message generated by Exim is necessarily in response
4016 to another message, thes have all been changed to:
4018 Auto-Submitted: auto-replied
4020 in accordance with these statements in the RFC:
4022 The auto-replied keyword:
4024 - SHOULD be used on messages sent in direct response to another
4025 message by an automatic process,
4027 - MUST NOT be used on manually-generated messages,
4029 - MAY be used on Delivery Status Notifications (DSNs) and Message
4030 Disposition Notifications (MDNs),
4032 - MUST NOT be used on messages generated by automatic or periodic
4033 processes, except for messages which are automatic responses to
4036 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4037 to the default Received: header definition.
4039 PH/12 Added log selector acl_warn_skipped (default on).
4041 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4042 variables because (a) they are in the wrong storage pool and (b) even if
4043 they were copied, it wouldn't work properly because of the caching.
4045 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4046 checking when decoding. Apparently there are clients that generate
4047 overlong encoded strings. Why am I not surprised?
4049 PH/15 If the first argument of "${if match_address" was not empty, but did not
4050 contain an "@" character, Exim crashed. Now it writes a panic log message
4051 and treats the condition as false.
4053 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4055 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4056 extension "envelope-auth". The code is finished and in agreement with
4057 other implementations, but there is no documentation so far and in fact,
4058 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4059 not changing the active code.
4061 Print executed "if" and "elsif" statements when debugging is used. This
4062 helps a great deal to understand what a filter does.
4064 Document more things not specified clearly in RFC3028. I had all this
4065 sorted out, when out of a sudden new issues came to my mind. Oops."
4067 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4070 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4071 canonical form (as documented). However, after a host name lookup from
4072 the IP address, check_host() was doing a simple string comparison with
4073 addresses acquired from the DNS when checking that the found name did
4074 have the original IP as one of its addresses. Since any found IPv6
4075 addresses are likely to be in abbreviated form, the comparison could
4076 fail. Luckily, there already exists a function for doing the comparison
4077 by converting both addresses to binary, so now that is used instead of
4078 the text comparison.
4080 PH/20 There was another similar case to PH/19, when a complete host name was
4081 given in a host list; looking up its IP address could give an abbreviated
4082 form, whereas the current host's name might or might not be abbreviated.
4083 The same fix has been applied.
4089 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4090 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4093 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4094 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4096 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4098 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4099 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4100 Also, for "Auto-Submitted", treat anything other than "no" as
4101 non-personal, in accordance with RFC 3834. (Previously it treated
4102 anything starting "auto-" as non-personal.)
4104 TF/01 The control=submission/name=... option had a problem with syntax
4105 errors if the name included a slash character. The /name= option
4106 now slurps the rest of the string, so it can include any characters
4107 but it must come last in the list of options (after /sender_retain
4110 PH/05 Some modifications to the interface to the fake nameserver for the new
4118 TK/01 Added the "success_on_redirect" address verification option. See
4119 NewStuff for rationale and an example.
4121 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4123 PH/02 Patch to exigrep to allow it to work on syslog lines.
4125 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4126 fread() to skip over the body file's header line, because in Cygwin the
4127 header line is locked and is inaccessible.
4129 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4130 co-exist for some time) to make it clear that it is the Exim ID that is
4131 referenced, not the Message-ID: header line.
4133 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4134 string_format() function, because snprintf() does not exist on all
4137 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4138 ${stat: expansion item.
4140 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4141 protocol synchronization error", to keep the pedants happy.
4143 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4144 well as for IRIX systems, when gcc is being used. See the host.c source
4147 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4149 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4152 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4153 transport and to the smtp transport in LMTP mode.
4155 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4157 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4158 run a filter in a subprocess. This could lead to confusion in subsequent
4159 lookups in the parent process. There should also be a search_tidyup() at
4160 the end of the subprocess.
4162 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4163 only if the host matched helo_try_verify_hosts, which caused the
4164 verification to occur when the EHLO/HELO command was issued. The ACL just
4165 tested the remembered result. Now, if a previous verification attempt has
4166 not happened, "verify = helo" does it there and then.
4168 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4170 TK/03 Fix log output including CR from clamd.
4172 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4173 exist provoked a memory error which could cause a segfault.
4175 PH/15 Installed PCRE 6.2
4177 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4179 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4180 of the problem. Specifically, suggested +O2 rather than +O1 for the
4183 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4185 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4186 record, the port was not correctly logged when the outgoing_port log
4187 selector was set (it logged the transort's default port).
4189 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4190 fallback_hosts, and "hosts" in the smtp transport.
4192 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4193 host errors such as "Connection refused".
4195 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4196 authentication with radiusclient 0.4.9:
4198 - Error returned from rc_read_config was caught wrongly
4199 - Username/password not passed on to radius server due to wrong length.
4201 The presumption is that some radiusclient API changes for 4.51/PH/17
4202 were not taken care of correctly. The code is still untested by me (my
4203 Linux distribution still has 0.3.2 of radiusclient), but it was
4204 contributed by a Radius user.
4206 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4207 expanding the "port" option of the smtp transport.
4209 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4210 while reading a MIME header. Thanks to Tom Hughes for a patch.
4212 PH/24 Include config.h inside local_scan.h so that configuration settings are
4215 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4216 that in an ACL for RCPT (for example), you can examine exactly what was
4219 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4220 commands, but it was not correctly comparing the address with the actual
4221 client host address. Thus, it would show the EHLO address in Received:
4222 header lines when this was not necessary.
4224 PH/27 Added the % operator to ${eval:}.
4226 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4227 it should be ignoring failures (because with -C, for example, it has lost
4228 privilege). It wasn't ignoring creation failures other than "already
4231 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4234 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4235 because an input line was too long, either on its own, or by virtue of
4236 too many continuations, the temporary file was not being removed, and the
4237 return code was incorrect.
4239 PH/31 Missing "BOOL" in function definition in filtertest.c.
4241 PH/32 Applied Sieve patches from the maintainer.
4243 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4245 PH/33 Added "verify = not_blind".
4247 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4248 Local/Makefile (with some defaults set). These are used in built scripts
4249 such as exicyclog, but they have never been used in the exim_install
4250 script (though there are many overriding facilities there). I have
4251 arranged that the exim_install script now takes note of these two
4254 PH/35 Installed configuration files for Dragonfly.
4256 PH/36 When a locally submitted message by a trusted user did not contain a
4257 From: header, and the sender address was obtained from -f or from an SMTP
4258 MAIL command, and the trusted user did not use -F to supply a sender
4259 name, $originator_name was incorrectly used when constructing a From:
4260 header. Furthermore, $originator_name was used for submission mode
4261 messages from external hosts without From: headers in a similar way,
4262 which is clearly wrong.
4264 PH/37 Added control=suppress_local_fixups.
4266 PH/38 When log_selector = +received_sender was set, and the addition of the
4267 sender made the log line's construction buffer exactly full, or one byte
4268 less than full, an overflow happened when the terminating "\n" was
4271 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4272 when the result of a list match is failure because a DNS lookup failed.
4274 PH/40 RM_COMMAND is now used in the building process.
4276 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4277 the "build-* directories that it finds.
4279 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4280 address in a domain literal was a prefix of an interface address.
4282 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4283 when verifying a sender address, unless rewrite_headers is false.
4285 PH/44 Wrote a long comment about why errors_to addresses are verified as
4286 recipients, not senders.
4288 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4289 the ratelimit ACL was added.
4291 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4293 PH/46 Added extra information about PostgreSQL errors to the error string.
4295 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4296 avoiding the need to install special zones in a real server. This is
4297 backwards compatible; if it can't find the fake resolver, it drops back.
4298 Thus, both old and new test suites can be run.
4300 TF/02 Added util/ratelimit.pl
4302 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4303 clock is set back in time.
4305 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4306 Candler <B.Candler@pobox.com>.
4308 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4309 OK for addresses that are the result of redirections.
4311 PH/48 A number of further additions for the benefit of the new test suite,
4312 including a fake gethostbyname() that interfaces to the fake DNS resolver
4315 TF/06 The fix for widen_domains has also been applied to qualify_single and
4316 search_parents which are the other dnslookup options that can cause
4319 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4322 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4324 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4325 service" if it encountered an MX record with an empty target hostname.
4326 The message is now "an MX or SRV record indicated no SMTP service".
4328 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4329 if the DNS of the sending site is misconfigured. This is quite a
4330 common situation. This change restores the behaviour of treating a
4331 helo verification defer as a failure.
4333 PH/51 If self=fail was set on a router, the bounce message did not include the
4334 actual error message.
4340 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4342 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4343 command itself ended in error, the underlying message about the transport
4344 filter timeout was being overwritten with the pipe command error. Now the
4345 underlying error message should be appended to the second error message.
4347 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4349 PH/02 Reduce the amount of output that "make" produces by default. Full output
4350 can still be requested.
4352 PH/03 The warning log line about a condition test deferring for a "warn" verb
4353 was being output only once per connection, rather than after each
4354 occurrence (because it was using the same function as for successful
4355 "warn" verbs). This seems wrong, so I have changed it.
4357 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4358 it should not have, which might have caused a crash in the right
4359 circumstances, but probably never did.
4361 PH/04 Installed a modified version of Tony Finch's patch to make submission
4362 mode fix the return path as well as the Sender: header line, and to
4363 add a /name= option so that you can make the user's friendly name appear
4366 TF/03 Added the control = fakedefer ACL modifier.
4368 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4369 Mark Lowes for thorough testing.
4371 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4373 TK/03 Merged latest SRS patch from Miles Wilton.
4375 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4376 with the definition in sysexits.h (which is #included earlier).
4377 Fortunately, Exim does not actually use EX_OK. The code used to try to
4378 preserve the sysexits.h value, by assuming that macro definitions were
4379 scanned for macro replacements. I have been disabused of this notion,
4380 so now the code just undefines EX_OK before #including unistd.h.
4382 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4383 in the smtp transport. When a block could not be written in a single
4384 write() function, the timeout was being re-applied to each part-write.
4385 This seems wrong - if the receiver was accepting one byte at a time it
4386 would take for ever. The timeout is now adjusted when this happens. It
4387 doesn't have to be particularly precise.
4389 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4390 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4392 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4393 without a domain if the check to <postmaster@domain> fails.
4395 SC/01 Eximstats: added -xls and the ability to specify output files
4396 (patch written by Frank Heydlauf).
4398 SC/02 Eximstats: use FileHandles for outputting results.
4400 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4402 SC/04 Eximstats: fixed display of large numbers with -nvr option
4404 SC/05 Eximstats: fixed merging of reports with empty tables.
4406 SC/06 Eximstats: added the -include_original_destination flag
4408 SC/07 Eximstats: removed tabs and trailing whitespace.
4410 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4412 TK/06 MBOX spool code: Add real "From " MBOX separator line
4413 so the .eml file is really in mbox format (even though
4414 most programs do not really care). Patch from Alex Miller.
4416 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4417 The latter is generated from $received_to and is only set if the
4418 message has one envelope recipient. SA can use these headers,
4419 obviously out-of-the-box. Patch from Alex Miller.
4421 PH/08 The ${def test on a variable was returning false if the variable's
4422 value was "0", contrary to what the specification has always said!
4423 The result should be true unless the variable is empty.
4425 PH/09 The syntax error of a character other than { following "${if
4426 def:variable_name" (after optional whitespace) was not being diagnosed.
4427 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4428 accidental colon was present, for example, could give incorrect results.
4430 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4431 result is used (not including appendfile, where other changes are about
4434 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4435 This involved changing a lot of size variables from int to off_t. It
4436 should work with maildirs and everything.
4438 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4439 spamd dying while we are connected to it.
4441 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4444 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4445 function for BDB 4.3.
4447 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4449 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4450 the sources for PCRE have been re-arranged and more files are now
4453 PH/15 The code I had for printing potentially long long variables in PH/11
4454 above was not the best (it lost precision). The length of off_t variables
4455 is now inspected at build time, and an appropriate printing format (%ld
4456 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4457 to be "long long int" or "long int". This is needed for the internal
4458 formatting function string_vformat().
4460 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4461 the configuration file to be ":syslog", then the script "guesses" where
4462 the logs files are, rather than using the compiled in default. In our
4463 case the guess is not the same as the compiled default, so the script
4464 suddenly stopped working when I started to use syslog. The patch checks
4465 to see if log_file_path is "". If so, it attempts to read it from exim
4466 with no configuration file to get the compiled in version, before it
4467 falls back to the previous guessing code."
4469 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4470 implementing BATV in an Exim configuration. See NewStuff for the gory
4473 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4474 Makefile that are specific to HP-UX.
4476 PH/18 If the "use_postmaster" option was set for a recipient callout together
4477 with the "random" option, the postmaster address was used as the MAIL
4478 FROM address for the random test, but not for the subsequent recipient
4479 test. It is now used for both.
4481 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4482 patch removes a few documentation additions to RFC 3028, because the
4483 latest draft now contains them. It adds the new en;ascii-case comparator
4484 and a new error check for 8bit text in MIME parts. Comparator and
4485 require names are now matched exactly. I enabled the subaddress
4486 extension, but it is not well tested yet (read: it works for me)."
4488 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4489 rework some of the code of TK/09 above to avoid the hardwired use of
4490 "%lld" and "long long". Replaced the call to snprintf() with a call to
4493 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4494 records point to non-existent hosts", "retry timeout exceeded", and
4495 "retry time not reached for any host after a long failure period".
4497 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4498 experimental DomainKeys support:
4500 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4501 (2) On an error such as an illegally used "control", the wrong name for
4502 the control was given.
4504 These problems did NOT occur unless DomainKeys support was compiled.
4506 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4508 PH/24 Added ${if match_ip condition.
4510 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4511 databases so that it will be absolutely obvious if a crash occurs in the
4512 DB library. This is a regular occurrence (often caused by mis-matched
4515 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4516 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4517 picked up on a user's system that detects such things. There doesn't seem
4518 to be a gcc warning option for this - only an attribute that has to be
4519 put on the function's prototype. It seems that in Fedora Core 4 they have
4520 set this on a number of new functions. No doubt there will be more in due
4523 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4524 specify a transport. However, if an address that was verified by such a
4525 router was the subject of a callout, Exim crashed because it tried to
4526 read the rcpt_include_affixes from the non-existent transport. Now it
4527 just assumes that the setting of that option is false. This bug was
4528 introduced by 4.51/PH/31.
4530 PH/28 Changed -d+all to exclude +memory, because that information is very
4531 rarely of interest, but it makes the output a lot bigger. People tend to
4532 do -d+all out of habit.
4534 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4535 code in os-type was giving problems when libc.so lives in lib64, like on
4538 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4539 aren't the modern standard, and it seems that some systems' include files
4540 don't always have them. Exim was already checking for some of the newer
4541 ones like T_AAAA, and defining it itself. I've added checks for all the
4542 record types that Exim uses.
4544 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4545 not automatically generating a new one, as it is supposed to. This
4546 prevented TLS from working. If the file did exist, but contained invalid
4547 data, a new version was generated, as expected. It was only the case of a
4548 non-existent file that was broken.
4550 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4551 with a change in libdomainkeys > 0.64.
4553 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4554 from DNS. If the selector record carries the flag, it now has
4555 precedence over the domain-wide flag.
4557 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4559 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4560 the use of a port name that isn't defined in /etc/services) occurred, the
4561 message was deferred as in a normal delivery, and thus remained on the
4562 spool, instead of being failed because of the mua_wrapper setting. This
4563 is now fixed, and I tidied up some of the mua_wrapper messages at the
4566 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4567 the messages in a hash of arrays rather than using individual hashes.
4568 This is a bit cleaner and results in dramatic memory savings, albeit
4569 at a slight CPU cost.
4571 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4572 as requested by Marc Sherman.
4574 SC/10 Eximstats: added histograms for user specified patterns as requested
4577 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4579 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4580 fopen() in the content-scanning modules that did not already have it.
4586 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4587 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4589 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4591 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4593 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4594 to test Sieve filters that use "vacation".
4596 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4597 that changes the way the GnuTLS parameters are stored in the cache file.
4598 The new format can be generated externally. For backward compatibility,
4599 if the data in the cache doesn't make sense, Exim assumes it has read an
4600 old-format file, and it generates new data and writes a new file. This
4601 means that you can't go back to an older release without removing the
4604 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4605 work if there are any delivery delays because "one_time" forces the
4606 parent to be marked "delivered", so its unseen clone is never tried
4607 again. For this reason, Exim now forbids the simultaneous setting of
4610 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4611 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4612 entry if you want to know the details. The fix, however, neglected to
4613 consider the case when local delivery batching is involved. The test for
4614 "previously delivered" was not happening when checking to see if an
4615 address could be batched with a previous (undelivered) one; under
4616 certain circumstances this could lead to multiple deliveries to the same
4619 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4620 in its include files, and this causes problems building Exim.
4622 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4623 header_syntax) but Exim was just ignoring anything given after a slash.
4624 In particular, this caused confusion with an attempt to use "verify =
4625 reverse_host_lookup/defer_ok". An error is now given when options are
4626 supplied for verify items that do not have them. (Maybe reverse_host_
4627 lookup should have a defer_ok option, but that's a different point.)
4629 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4630 defined by RFC 821) to 2048, because there were problems with some AUTH
4631 commands, and RFC 1869 says the size should be increased for extended
4632 SMTP commands that take arguments.
4634 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4637 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4638 "unknown" error; now it says that the functionality isn't in the binary.
4640 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4641 an address' error message when a string expansion fails (syntax or
4642 whatever). Otherwise the password may appear in the log. Following change
4643 PH/42 below, there is no longer a chance of it appearing in a bounce
4646 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4648 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4649 of its addresses was ever tried. (Bugzilla bug #2.)
4651 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4652 the result incorrectly in the debug output. (It correctly added a newline
4653 to what was transported.)
4655 TF/01 Added $received_time.
4657 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4658 commented out examples of how to interface to a virus scanner and to
4659 SpamAssassin. Also added commented examples of av_scanner and
4660 spamd_address settings.
4662 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4663 and controls are allowed in which ACLs. There were a couple of minor
4664 errors. Some of the entries in the conditions table (which is a table of
4665 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4666 negation of where the condition IS allowed.
4668 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4670 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4671 header file does not have a version number, so I've had to invent a new
4672 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4673 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4674 radiusclient), but it was contributed by a Radius user.
4676 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4677 files or directories, for OpenSSL.
4679 PH/19 When an Exim process that is running as root has to create an Exim log
4680 file, it does so in a subprocess that runs as exim:exim so as to get the
4681 ownership right at creation (otherwise, other Exim processes might see
4682 the file with the wrong ownership). There was no test for failure of this
4683 fork() call, which would lead to the process getting stuck as it waited
4684 for a non-existent subprocess. Forks do occasionally fail when resources
4685 run out. I reviewed all the other calls to fork(); they all seem to check
4688 PH/20 When checking for unexpected SMTP input at connect time (before writing
4689 the banner), Exim was not dealing correctly with a non-positive return
4690 from the read() function. If the client had disconnected by this time,
4691 the result was a log entry for a synchronization error with an empty
4692 string after "input=" when read() returned zero. If read() returned -1
4693 (an event I could not check), uninitialized data bytes were printed.
4694 There were reports of junk text (parts of files, etc) appearing after
4697 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4699 PH/22 Added support for macro redefinition, and (re)definition in between
4700 driver and ACL definitions.
4702 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4703 forgetting to use the resulting value; it was using the unexpanded value.
4705 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4706 hadn't been configured. The fix is from Juergen Kreileder, who
4707 understands it better than I do:
4709 "Here's what I see happening with three configured cyrus_sasl
4710 authenticators configured (plain, login, cram-md5):
4712 On startup auth_cyrus_sasl_init() gets called for each of these.
4713 This means three calls to sasl_listmech() without a specified mech_list.
4714 => SASL tests which mechs of all available mechs actually work
4715 => three warnings about OTP not working
4716 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4718 With the patch, sasl_listmech() also gets called three times. But now
4719 SASL's mech_list option is set to the server_mech specified in the the
4720 authenticator. Or in other words, the answer from sasl_listmech()
4721 gets limited to just the mech you're testing for (which is different
4723 => the return list contains just 'plain' or 'login', 'cram-md5' or
4724 nothing depending on the value of ob->server_mech.
4726 I've just tested the patch: Authentication still works fine,
4727 unavailable mechs specified in the exim configuration are still
4728 caught, and the auth.log warnings about OTP are gone."
4730 PH/25 When debugging is enabled, the contents of the command line are added
4731 to the debugging output, even when log_selector=+arguments is not
4734 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4735 answer is "GNU", and only if the return is "GNU/something" is the answer
4738 PH/27 $acl_verify_message is now set immediately after the failure of a
4739 verification in an ACL, and so is available in subsequent modifiers. In
4740 particular, the message can be preserved by coding like this:
4742 warn !verify = sender
4743 set acl_m0 = $acl_verify_message
4745 Previously, $acl_verify_message was set only while expanding "message"
4746 and "log_message" when a very denied access.
4748 PH/28 Modified OS/os.c-Linux with
4750 -#ifndef OS_LOAD_AVERAGE
4751 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4753 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4754 nomenclature these days.)
4756 PH/29 Installed patch from the Sieve maintainer that adds the options
4757 sieve_useraddress and sieve_subaddress to the redirect router.
4759 PH/30 In these circumstances:
4760 . Two addresses routed to the same list of hosts;
4761 . First host does not offer TLS;
4762 . First host accepts first address;
4763 . First host gives temporary error to second address;
4764 . Second host offers TLS and a TLS session is established;
4765 . Second host accepts second address.
4766 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4767 and peerdn, if requested) that were in fact used only for the second
4770 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4771 attention to any local part prefix or suffix that was matched by the
4772 router that accepted the address. It now behaves in the same way as it
4773 does for delivery: the affixes are removed from the local part unless
4774 rcpt_include_affixes is set on the transport.
4776 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4777 timeout during the DATA phase of an incoming message.
4779 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4780 applied a patch sanctioned by the Sieve maintainer.
4782 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4783 the uid or gid is negative. A case of a negative gid caused this to be
4784 noticed. The fix allows for either to be negative.
4786 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4787 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4790 PH/36 The change PH/12 above was broken. Fixed it.
4792 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4793 the grounds that routing the same address twice would always produce the
4794 same answer. This might have been true once, but it is certainly no
4795 longer true now. Routing a child address may depend on the previous
4796 routing that produced that child. Some complicated redirection strategies
4797 went wrong when messages had multiple recipients, and made Exim's
4798 behaviour dependent on the order in which the addresses were given.
4800 I have moved the duplicate checking until after the routing is complete.
4801 Exim scans the addresses that are assigned to local and remote
4802 transports, and removes any duplicates. This means that more work will be
4803 done, as duplicates will always all be routed, but duplicates are
4804 presumably rare, so I don't expect this is of any significance.
4806 For deliveries to pipes, files, and autoreplies, the duplicate checking
4807 still happens during the routing process, since they are not going to be
4810 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4811 It corrects a timeout issue with spamd. This is Ian's comment: "The
4812 background is that sometimes spamd either never reads data from a
4813 connection it has accepted, or it never writes response data. The exiscan
4814 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4815 blindly assumes that writes won't block so it may never time out."
4817 PH/39 Allow G after quota size as well as K and M.
4819 PH/40 The value set for $authenticated_id in an authenticator may not contain
4820 binary zeroes or newlines because the value is written to log lines and
4821 to spool files. There was no check on this. Now the value is run through
4822 the string_printing() function so that such characters are converted to
4823 printable escape sequences.
4825 PH/41 $message_linecount is a new variable that contains the total number of
4826 lines in the message. Compare $body_linecount, which is the count for the
4829 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4830 bounce and delay warning messages, except in certain special cases, which
4833 (a) An SMTP error message from a remote host;
4834 (b) A message specified in a :fail: redirection;
4835 (c) A message specified in a "fail" command in a system filter;
4836 (d) A message specified in a FAIL return from the queryprogram router;
4837 (e) A message specified by the cannot_route_message router option.
4839 In these cases only, Exim does include the error details in bounce and
4840 warning messages. There are also a few cases where bland messages such
4841 as "unrouteable address" or "local delivery error" are given.
4843 PH/43 $value is now also set for the "else" part of a ${run expansion.
4845 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4846 being worked on, but at least Exim now implements the latest version to
4849 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4850 process to complete was treated as a delivery failure, a timeout while
4851 writing the message to the pipe was logged, but erroneously treated as a
4852 successful delivery. Such timeouts include transport filter timeouts. For
4853 consistency with the overall process timeout, these timeouts are now
4854 treated as errors, giving rise to delivery failures by default. However,
4855 there is now a new Boolean option for the pipe transport called
4856 timeout_defer, which, if set TRUE, converts the failures into defers for
4857 both kinds of timeout. A transport filter timeout is now identified in
4860 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4861 systems where "make" and "gmake" are different, calling "gmake" at top
4862 level broke things. I've arranged for the value of $(MAKE) to be passed
4863 from the Makefile to this script so that it can call the same version of
4867 A note about Exim versions 4.44 and 4.50
4868 ----------------------------------------
4870 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4871 changes of various kinds. As a consequence, a big documentation update was
4872 needed. This delayed the release for rather longer than seemed good, especially
4873 in the light of a couple of (minor) security issues. Therefore, the changes
4874 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4875 release. So 4.44 and 4.50 are in effect two different branches that both start
4878 I have left the 4.50 change log unchanged; it contains all the changes since
4879 4.43. The change log for 4.44 is below; many of its items are identical to
4880 those for 4.50. This seems to be the most sensible way to preserve the
4881 historical information.
4887 1. Minor wording change to the doc/README.SIEVE file.
4889 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4890 computation of the current number of files was incorrect.
4892 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4893 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4896 4. Give more explanation in the error message when the command for a transport
4897 filter fails to execute.
4899 5. There are several places where Exim runs a non-Exim command in a
4900 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4901 was being done only for the command run by the queryprogram router. It is
4902 now done for all such subprocesses. The other cases are: ${run, transport
4903 filters, and the commands run by the lmtp and pipe transports.
4905 6. Added CONFIGURE_GROUP build-time option.
4907 7. Some older OS have a limit of 256 on the maximum number of file
4908 descriptors. Exim was using setrlimit() to set 1000 as a large value
4909 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4910 systems. I've change it so that if it can't get 1000, it tries for 256.
4912 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4913 was an oversight, and furthermore, ever since the addition of extra
4914 controls (e.g. 4.43/32), the checks on when to allow different forms of
4915 "control" were broken. There should now be diagnostics for all cases when a
4916 control that does not make sense is encountered.
4918 9. Added the /retain_sender option to "control=submission".
4920 10. $recipients is now available in the predata ACL (oversight).
4922 11. Tidy the search cache before the fork to do a delivery from a message
4923 received from the command line. Otherwise the child will trigger a lookup
4924 failure and thereby defer the delivery if it tries to use (for example) a
4925 cached ldap connection that the parent has called unbind on.
4927 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4928 of $address_data from the recipient verification was clobbered by the
4929 sender verification.
4931 13. The value of address_data from a sender verification is now available in
4932 $sender_address_data in subsequent conditions in the ACL statement.
4934 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4936 15. Added a new option "connect=<time>" to callout options, to set a different
4939 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4940 was its contents. (It was OK if the option was not defined at all.)
4942 17. A "Completed" log line is now written for messages that are removed from
4943 the spool by the -Mrm option.
4945 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4946 information about exactly what failed.
4948 19. Added -dd to debug only the daemon process.
4950 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4951 handles timeouts, both on the server side and network timeouts. Renamed the
4952 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4954 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4955 It is now set to "smtps".
4957 22. $host_address is now set to the target address during the checking of
4958 ignore_target_hosts.
4960 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4961 being passed; this would have caused $sender_host_name to have been used if
4962 matching the list had actually called for a host name (not very likely,
4963 since this list is usually IP addresses). A host name is now passed as
4966 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4967 code that specifies a non-restarting handler (typically sigaction() in
4968 modern systems) in an attempt to fix a rare and obscure crash bug.
4970 25. Narrowed the window for a race in the daemon that could cause it to ignore
4971 SIGCHLD signals. This is not a major problem, because they are used only to
4972 wake it up if nothing else does.
4974 26. A malformed maildirsize file could cause Exim to calculate negative values
4975 for the mailbox size or file count. Odd effects could occur as a result.
4976 The maildirsize information is now recalculated if the size or filecount
4979 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4980 support for a long time. Removed HAVE_SYS_VFS_H.
4982 28. Installed the latest version of exipick from John Jetmore.
4984 29. In an address list, if the pattern was not a regular expression, an empty
4985 subject address (from a bounce message) matched only if the pattern was an
4986 empty string. Non-empty patterns were not even tested. This was the wrong
4987 because it is perfectly reasonable to use an empty address as part of a
4988 database query. An empty address is now tested by patterns that are
4989 lookups. However, all the other forms of pattern expect the subject to
4990 contain a local part and a domain, and therefore, for them, an empty
4991 address still always fails if the pattern is not itself empty.
4993 30. Exim went into a mad DNS loop when attempting to do a callout where the
4994 host was specified on an smtp transport, and looking it up yielded more
4995 than one IP address.
4997 31. Re-factored the code for checking spool and log partition space into a
4998 function that finds that data and another that does the check. The former
4999 is then used to implement four new variables: $spool_space, $log_space,
5000 $spool_inodes, and $log_inodes.
5002 32. The RFC2047 encoding function was originally intended for short strings
5003 such as real names; it was not keeping to the 75-character limit for
5004 encoded words that the RFC imposes. It now respects the limit, and
5005 generates multiple encoded words if necessary. To be on the safe side, I
5006 have increased the buffer size for the ${rfc2047: expansion operator from
5009 33. It is now permitted to omit both strings after an "if" condition; if the
5010 condition is true, the result is "true". As before, when the second string
5011 is omitted, a false condition yields an empty string. This makes it less
5012 cumbersome to write custom ACL and router conditions.
5014 34. Failure to deliver a bounce message always caused it to be frozen, even if
5015 there was an errors_to setting on the router. The errors_to setting is now
5018 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5019 canonical form (fully expanded) before being placed in
5020 $sender_host_address.
5022 36. The table in the code that translates DNS record types into text (T_A to
5023 "A" for instance) was missing entries for NS and CNAME. It is just possible
5024 that this could have caused confusion if both these types were looked up
5025 for the same domain, because the text type is used as part of Exim's
5026 per-process caching. But the chance of anyone hitting this buglet seems
5029 37. The dnsdb lookup has been extended in a number of ways.
5031 (1) There is a new type, "zns", which walks up the domain tree until it
5032 finds some nameserver records. It should be used with care.
5034 (2) There is a new type, "mxh", which is like "mx" except that it returns
5035 just the host names, not the priorities.
5037 (3) It is now possible to give a list of domains (or IP addresses) to be
5038 looked up. The behaviour when one of the lookups defers can be
5039 controlled by a keyword.
5041 (4) It is now possible to specify the separator character for use when
5042 multiple records are returned.
5044 38. The dnslists ACL condition has been extended: it is now possible to supply
5045 a list of IP addresses and/or domains to be looked up in a particular DNS
5048 39. Added log_selector=+queue_time_overall.
5050 40. When running the queue in the test harness, wait just a tad after forking a
5051 delivery process, to get repeatability of debugging output.
5053 41. Include certificate and key file names in error message when GnuTLS fails
5054 to set them up, because the GnuTLS error message doesn't include the name
5055 of the failing file when there is a problem reading it.
5057 42. Allow both -bf and -bF in the same test run.
5059 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5061 44. The "Exiscan patch" is now merged into the mainline Exim source.
5063 45. Sometimes the final signoff response after QUIT could fail to get
5064 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5065 < 0 before doing a fflush(). This bug looks as though it goes back to the
5066 introduction of TLS in release 3.20, but "sometimes" must have been rare
5067 because the tests only now provoked it.
5069 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5070 (this can affect the format of dates).
5072 47. exim_tidydb, when checking for the continued existence of a message for
5073 which it has found a message-specific retry record, was not finding
5074 messages that were in split spool directories. Consequently, it was
5075 deleting retry records that should have stayed in existence.
5077 48. Steve fixed some bugs in eximstats.
5079 49. The SPA authentication driver was not abandoning authentication and moving
5080 on to the next authenticator when an expansion was forced to fail,
5081 contradicting the general specification for all authenticators. Instead it
5082 was generating a temporary error. It now behaves as specified.
5084 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5085 (the order specifies the preference for clients). The order is now AES256,
5086 AES128, 3DES, ARCFOUR128.
5088 51. Small patch to Sieve code - explicitly set From: when generating an
5091 52. Exim crashed if a remote delivery caused a very long error message to be
5092 recorded - for instance if somebody sent an entire SpamAssassin report back
5093 as a large number of 550 error lines. This bug was coincidentally fixed by
5094 increasing the size of one of Exim's internal buffers (big_buffer) that
5095 happened as part of the Exiscan merge. However, to be on the safe side, I
5096 have made the code more robust (and fixed the comments that describe what
5099 53. Now that there can be additional text after "Completed" in log lines (if
5100 the queue_time_overall log selector is set), a one-byte patch to exigrep
5101 was needed to allow it to recognize "Completed" as not the last thing in
5104 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5105 patch that reportedly fixes this has been added. I am not expert enough to
5106 create a test for it. This is what the patch creator wrote:
5108 "I found a little strange behaviour of ldap code when working with
5109 Windows 2003 AD Domain, where users was placed in more than one
5110 Organization Units. When I tried to give exim partial DN, the exit code
5111 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5112 But simultaneously result of request was absolutely normal ldap result,
5113 so I produce this patch..."
5115 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5116 REFERENCE, so I have modified the code to exclude the patch when that macro
5119 55. Some experimental protocols are using DNS PTR records for new purposes. The
5120 keys for these records are domain names, not reversed IP addresses. The
5121 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5122 leaves it alone. Component reversal etc. now happens only for IP addresses.
5125 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5127 57. Double the size of the debug message buffer (to 2048) so that more of very
5128 long debug lines gets shown.
5130 58. The exicyclog utility now does better if the number of log files to keep
5131 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5133 59. Two changes related to the smtp_active_hostname option:
5135 (1) $smtp_active_hostname is now available as a variable.
5136 (2) The default for smtp_banner uses $smtp_active_hostname instead
5137 of $primary_hostname.
5139 60. The host_aton() function is supposed to be passed a string that is known
5140 to be a valid IP address. However, in the case of IPv6 addresses, it was
5141 not checking this. This is a hostage to fortune. Exim now panics and dies
5142 if the condition is not met. A case was found where this could be provoked
5143 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5144 components; fortuitously, this particular loophole had already been fixed
5145 by change 4.50/55 above.
5147 If there are any other similar loopholes, the new check in host_aton()
5148 itself should stop them being exploited. The report I received stated that
5149 data on the command line could provoke the exploit when Exim was running as
5150 exim, but did not say which command line option was involved. All I could
5151 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5152 running as the user.
5155 61. There was a buffer overflow vulnerability in the SPA authentication code
5156 (which came originally from the Samba project). I have added a test to the
5157 spa_base64_to_bits() function which I hope fixes it.
5160 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5161 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5163 63. The daemon start-up calls getloadavg() while still root for those OS that
5164 need the first call to be done as root, but it missed one case: when
5165 deliver_queue_load_max is set with deliver_drop_privilege. This is
5166 necessary for the benefit of the queue runner, because there is no re-exec
5167 when deliver_drop_privilege is set.
5169 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5170 This has been fixed.
5172 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5173 was in use, was not putting the data itself into the right store pool;
5174 consequently, it could be overwritten for a subsequent message in the same
5175 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5178 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5180 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5181 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5182 the function were treating the return as a boolean value, which happened to
5183 work because 0=false and not-0=true, but is not correct code.
5185 68. The host_aton() function was not handling scoped IPv6 addresses (those
5186 with, for example, "%eth0" on the end) correctly.
5188 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5189 negated items (that is, ~something) in unsigned ints. Some compilers
5190 apparently mutter when there is no cast.
5192 70. If an address verification called from an ACL failed, and did not produce a
5193 user-specific message (i.e. there was only a "system" message), nothing was
5194 put in $acl_verify_message. In this situation, it now puts the system
5197 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5198 session; change 4.31/43 added the unwanted input to the log line - except
5199 that it did not do this in the start of session case. It now does.
5201 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5202 This is wrong and can cause the other end to generate a synchronization
5203 error if it is another Exim or anything else that does the synchronization
5204 check. A QUIT command is no longer sent after a timeout.
5206 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5207 during host lookups.
5209 74. The defer_ok option of callout verification was not working if it was used
5210 when verifying addresses in header lines, that is, for this case:
5212 verify = header_sender/callout=defer_ok
5214 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5215 those file descriptors could be used for SMTP connections. If anything
5216 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5217 could be sent to the SMTP client, causing chaos. The daemon now opens
5218 stdin, stdout, and stderr to /dev/null when it puts itself into the
5221 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5222 log by default. The user can override this with suitable Perl magic.
5224 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5225 the log message when discard triggers, was not working for the DATA ACL or
5226 for the non-SMTP ACL.
5228 78. Error message wording change in sieve.c.
5230 79. If smtp_accept_max_per_host was set, the number of connections could be
5231 restricted to fewer than expected, because the daemon was trying to set up
5232 a new connection before checking whether the processes handling previous
5233 connections had finished. The check for completed processes is now done
5234 earlier. On busy systems, this bug wouldn't be noticed because something
5235 else would have woken the daemon, and it would have reaped the completed
5238 80. If a message was submitted locally by a user whose login name contained one
5239 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5240 It caused a spool format error. I have fixed the spool reading code. A
5241 related problem was that the "from" clause in the Received: line became
5242 illegal because of the space(s). It is now covered by ${quote_local_part.
5244 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5247 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5249 83. Patch from Sieve maintainer to fix unterminated string problem in
5250 "vacation" handling.
5252 84. Some minor changes to the Linux configuration files to help with other
5253 OS variants using glibc.
5255 85. One more patch for Sieve to update vacation handling to latest spec.
5258 ----------------------------------------------------
5259 See the note above about the 4.44 and 4.50 releases.
5260 ----------------------------------------------------
5266 1. Change 4.43/35 introduced a bug that caused file counts to be
5267 incorrectly computed when quota_filecount was set in an appendfile
5270 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5271 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5274 3. Give more explanation in the error message when the command for a transport
5275 filter fails to execute.
5277 4. There are several places where Exim runs a non-Exim command in a
5278 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5279 was being done only for the command run by the queryprogram router. It is
5280 now done for all such subprocesses. The other cases are: ${run, transport
5281 filters, and the commands run by the lmtp and pipe transports.
5283 5. Some older OS have a limit of 256 on the maximum number of file
5284 descriptors. Exim was using setrlimit() to set 1000 as a large value
5285 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5286 systems. I've change it so that if it can't get 1000, it tries for 256.
5288 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5289 was an oversight, and furthermore, ever since the addition of extra
5290 controls (e.g. 4.43/32), the checks on when to allow different forms of
5291 "control" were broken. There should now be diagnostics for all cases when a
5292 control that does not make sense is encountered.
5294 7. $recipients is now available in the predata ACL (oversight).
5296 8. Tidy the search cache before the fork to do a delivery from a message
5297 received from the command line. Otherwise the child will trigger a lookup
5298 failure and thereby defer the delivery if it tries to use (for example) a
5299 cached ldap connection that the parent has called unbind on.
5301 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5302 of $address_data from the recipient verification was clobbered by the
5303 sender verification.
5305 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5306 was its contents. (It was OK if the option was not defined at all.)
5308 11. A "Completed" log line is now written for messages that are removed from
5309 the spool by the -Mrm option.
5311 12. $host_address is now set to the target address during the checking of
5312 ignore_target_hosts.
5314 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5315 being passed; this would have caused $sender_host_name to have been used if
5316 matching the list had actually called for a host name (not very likely,
5317 since this list is usually IP addresses). A host name is now passed as
5320 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5321 code that specifies a non-restarting handler (typically sigaction() in
5322 modern systems) in an attempt to fix a rare and obscure crash bug.
5324 15. Narrowed the window for a race in the daemon that could cause it to ignore
5325 SIGCHLD signals. This is not a major problem, because they are used only to
5326 wake it up if nothing else does.
5328 16. A malformed maildirsize file could cause Exim to calculate negative values
5329 for the mailbox size or file count. Odd effects could occur as a result.
5330 The maildirsize information is now recalculated if the size or filecount
5333 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5334 support for a long time. Removed HAVE_SYS_VFS_H.
5336 18. Updated exipick to current release from John Jetmore.
5338 19. Allow an empty sender to be matched against a lookup in an address list.
5339 Previously the only cases considered were a regular expression, or an
5342 20. Exim went into a mad DNS lookup loop when doing a callout where the
5343 host was specified on the transport, if the DNS lookup yielded more than
5346 21. The RFC2047 encoding function was originally intended for short strings
5347 such as real names; it was not keeping to the 75-character limit for
5348 encoded words that the RFC imposes. It now respects the limit, and
5349 generates multiple encoded words if necessary. To be on the safe side, I
5350 have increased the buffer size for the ${rfc2047: expansion operator from
5353 22. Failure to deliver a bounce message always caused it to be frozen, even if
5354 there was an errors_to setting on the router. The errors_to setting is now
5357 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5358 canonical form (fully expanded) before being placed in
5359 $sender_host_address.
5361 24. Updated eximstats to version 1.33
5363 25. Include certificate and key file names in error message when GnuTLS fails
5364 to set them up, because the GnuTLS error message doesn't include the name
5365 of the failing file when there is a problem reading it.
5367 26. Expand error message when OpenSSL has problems setting up cert/key files.
5370 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5371 (this can affect the format of dates).
5373 28. exim_tidydb, when checking for the continued existence of a message for
5374 which it has found a message-specific retry record, was not finding
5375 messages that were in split spool directories. Consequently, it was
5376 deleting retry records that should have stayed in existence.
5378 29. eximstats updated to version 1.35
5379 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5380 1.35 - bugfix such that pie charts by volume are generated correctly
5382 30. The SPA authentication driver was not abandoning authentication and moving
5383 on to the next authenticator when an expansion was forced to fail,
5384 contradicting the general specification for all authenticators. Instead it
5385 was generating a temporary error. It now behaves as specified.
5387 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5388 (the order specifies the preference for clients). The order is now AES256,
5389 AES128, 3DES, ARCFOUR128.
5391 31. Small patch to Sieve code - explicitly set From: when generating an
5394 32. Exim crashed if a remote delivery caused a very long error message to be
5395 recorded - for instance if somebody sent an entire SpamAssassin report back
5396 as a large number of 550 error lines. This bug was coincidentally fixed by
5397 increasing the size of one of Exim's internal buffers (big_buffer) that
5398 happened as part of the Exiscan merge. However, to be on the safe side, I
5399 have made the code more robust (and fixed the comments that describe what
5402 33. Some experimental protocols are using DNS PTR records for new purposes. The
5403 keys for these records are domain names, not reversed IP addresses. The
5404 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5405 leaves it alone. Component reversal etc. now happens only for IP addresses.
5408 34. The host_aton() function is supposed to be passed a string that is known
5409 to be a valid IP address. However, in the case of IPv6 addresses, it was
5410 not checking this. This is a hostage to fortune. Exim now panics and dies
5411 if the condition is not met. A case was found where this could be provoked
5412 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5413 components; fortuitously, this particular loophole had already been fixed
5414 by change 4.50/55 or 4.44/33 above.
5416 If there are any other similar loopholes, the new check in host_aton()
5417 itself should stop them being exploited. The report I received stated that
5418 data on the command line could provoke the exploit when Exim was running as
5419 exim, but did not say which command line option was involved. All I could
5420 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5421 running as the user.
5424 35. There was a buffer overflow vulnerability in the SPA authentication code
5425 (which came originally from the Samba project). I have added a test to the
5426 spa_base64_to_bits() function which I hope fixes it.
5429 36. The daemon start-up calls getloadavg() while still root for those OS that
5430 need the first call to be done as root, but it missed one case: when
5431 deliver_queue_load_max is set with deliver_drop_privilege. This is
5432 necessary for the benefit of the queue runner, because there is no re-exec
5433 when deliver_drop_privilege is set.
5435 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5436 was in use, was not putting the data itself into the right store pool;
5437 consequently, it could be overwritten for a subsequent message in the same
5438 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5441 38. Sometimes the final signoff response after QUIT could fail to get
5442 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5443 < 0 before doing a fflush(). This bug looks as though it goes back to the
5444 introduction of TLS in release 3.20, but "sometimes" must have been rare
5445 because the tests only now provoked it.
5451 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5452 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5453 Now it returns an integer. A number of calls were still expecting a T/F
5454 return. Fortuitously, in all cases, the tests worked in OK situations,
5455 which is the norm. However, things would have gone wrong on any write
5456 failures on the smtp file descriptor. This function is used when sending
5457 messages over SMTP and also when doing verify callouts.
5459 2. When Exim is called to do synchronous delivery of a locally submitted
5460 message (the -odf or -odi options), it no longer closes stderr before doing
5463 3. Implemented the mua_wrapper option.
5465 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5467 5. Implemented the functions header_remove(), header_testname(),
5468 header_add_at_position(), and receive_remove_recipient(), and exported them
5471 6. If an ACL "warn" statement specified the addition of headers, Exim already
5472 inserted X-ACL-Warn: at the start if there was no header name. However, it
5473 was not making this test for the second and subsequent header lines if
5474 there were newlines in the string. This meant that an invalid header could
5475 be inserted if Exim was badly configured.
5477 7. Allow an ACL "warn" statement to add header lines at the start or after all
5478 the Received: headers, as well as at the end.
5480 8. Added the rcpt_4xx retry error code.
5482 9. Added postmaster_mailfrom=xxx to callout verification option.
5484 10. Added mailfrom=xxxx to the callout verification option, for verify=
5487 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5488 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5490 12. Inserted some casts to stop certain compilers warning when using pointer
5491 differences as field lengths or precisions in printf-type calls (mostly
5492 affecting debugging statements).
5494 13. Added optional readline() support for -be (dynamically loaded).
5496 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5497 same clock tick as a message's arrival, so that its received time was the
5498 same as the "first fail" time on the retry record, and that message
5499 remained on the queue past the ultimate address timeout, every queue runner
5500 would try a delivery (because it was past the ultimate address timeout) but
5501 after another failure, the ultimate address timeout, which should have then
5502 bounced the address, did not kick in. This was a "< instead of <=" error;
5503 in most cases the first failure would have been in the next clock tick
5504 after the received time, and all would be well.
5506 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5507 being recognized when the domain list was tested by the match_domain
5508 condition in an expansion string.
5510 16. Added the ${str2b64: operator.
5512 17. Exim was always calling setrlimit() to set a large limit for the number of
5513 processes, without checking whether the existing limit was already
5514 adequate. (It did check for the limit on file descriptors.) Furthermore,
5515 errors from getrlimit() and setrlimit() were being ignored. Now they are
5516 logged to the main and panic logs, but Exim does carry on, to try to do its
5517 job under whatever limits there are.
5519 18. Imported PCRE 5.0.
5521 19. Trivial typo in log message " temporarily refused connection" (the leading
5524 20. If the log selector return_path_on_delivery was set and an address was
5525 redirected to /dev/null, the delivery process crashed because it assumed
5526 that a return path would always be set for a "successful" delivery. In this
5527 case, the whole delivery is bypassed as an optimization, and therefore no
5530 21. Internal re-arrangement: the function for sending a challenge and reading
5531 a response while authentication was assuming a zero-terminated challenge
5532 string. It's now changed to take a pointer and a length, to allow for
5533 binary data in such strings.
5535 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5537 23. Exim was not respecting finduser_retries when seeking the login of the
5538 uid under which it was called; it was always trying 10 times. (The default
5539 setting of finduser_retries is zero.) Also, it was sleeping after the final
5540 failure, which is pointless.
5542 24. Implemented tls_on_connect_ports.
5544 25. Implemented acl_smtp_predata.
5546 26. If the domain in control=submission is set empty, Exim assumes that the
5547 authenticated id is a complete email address when it generates From: or
5548 Sender: header lines.
5550 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5551 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5552 chown and chgrp in /bin and hostname in /usr/bin.
5554 28. Exim was keeping the "process log" file open after each use, just as it
5555 does for the main log. This opens the possibility of it remaining open for
5556 long periods when the USR1 signal hits a daemon. Occasional processlog
5557 errors were reported, that could have been caused by this. Anyway, it seems
5558 much more sensible not to leave this file open at all, so that is what now
5561 29. The long-running daemon process does not normally write to the log once it
5562 has entered its main loop, and it closes the log before doing so. This is
5563 so that log files can straightforwardly be renamed and moved. However,
5564 there are a couple of unusual error situations where the daemon does write
5565 log entries, and I had neglected to close the log afterwards.
5567 30. The text of an SMTP error response that was received during a remote
5568 delivery was being truncated at 512 bytes. This is too short for some of
5569 the long messages that one sometimes sees. I've increased the limit to
5572 31. It is now possible to make retry rules that apply only when a message has a
5573 specific sender, in particular, an empty sender.
5575 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5576 it possible to be selective about when SMTP synchronization is enforced.
5578 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5580 32. Implemented hosts_connection_nolog.
5582 33. Added an ACL for QUIT.
5584 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5587 35. Added mailbox_size and mailbox_filecount to appendfile.
5589 36. Added control = no_multiline_responses to ACLs.
5591 37. There was a bug in the logic of the code that waits for the clock to tick
5592 in the case where the clock went backwards by a substantial amount such
5593 that the microsecond fraction of "now" was more than the microsecond
5594 fraction of "then" (but the whole seconds number was less).
5596 38. Added support for the libradius Radius client library this is found on
5597 FreeBSD (previously only the radiusclient library was supported).
5603 1. When certain lookups returned multiple values in the form name=value, the
5604 quoting of the values was not always being done properly. Specifically:
5605 (a) If the value started with a double quote, but contained no whitespace,
5607 (b) If the value contained whitespace other than a space character (i.e.
5608 tabs or newlines or carriage returns) it was not quoted.
5609 This fix has been applied to the mysql and pgsql lookups by writing a
5610 separate quoting function and calling it from the lookup code. The fix
5611 should probably also be applied to nisplus, ibase and oracle lookups, but
5612 since I cannot test any of those, I have not disturbed their existing code.
5614 2. A hit in the callout cache for a specific address caused a log line with no
5615 reason for rejecting RCPT. Now it says "Previous (cached) callout
5616 verification failure".
5618 3. There was an off-by-one bug in the queryprogram router. An over-long
5619 return line was truncated at 256 instead of 255 characters, thereby
5620 overflowing its buffer with the terminating zero. As well as fixing this, I
5621 have increased the buffer size to 1024 (and made a note to document this).
5623 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5624 when Exim is waiting for an SMTP response from a remote server, Exim
5625 restarts its select() call on the socket, thereby resetting its timeout.
5626 This is not a problem when such interrupts are rare. Somebody set up a cron
5627 job to run exiwhat every 2 minutes, which is less than the normal select()
5628 timeout (5 or 10 minutes). This meant that the select() timeout never
5629 kicked in because it was always reset. I have fixed this by comparing the
5630 time when an interrupt arrives with the time at the start of the first call
5631 to select(). If more time than the timeout has elapsed, the interrupt is
5632 treated as a timeout.
5634 5. Some internal re-factoring in preparation for the addition of Sieve
5635 extensions (by MH). In particular, the "personal" test is moved to a
5636 separate function, and given an option for scanning Cc: and Bcc: (which is
5637 not set for Exim filters).
5639 6. When Exim created an email address using the login of the caller as the
5640 local part (e.g. when creating a From: or Sender: header line), it was not
5641 quoting the local part when it contained special characters such as @.
5643 7. Installed new OpenBSD configuration files.
5645 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5646 try to make them clearer.
5648 9. Callout options, other than the timeout value, were being ignored when
5649 verifying sender addresses in header lines. For example, when using
5651 verify = header_sender/callout=no_cache
5653 the cache was (incorrectly) being used.
5655 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5656 only the Cygwin environment.
5658 11. When return_path_on_delivery was set as a log selector, if different remote
5659 addresses in the same message used different return paths and parallel
5660 remote delivery occurred, the wrong values would sometimes be logged.
5661 (Whenever a remote delivery process finished, the return path value from
5662 the most recently started remote delivery process was logged.)
5664 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5665 lines when AUTH and/or TLS are in use. This is the "received protocol"
5666 field. Exim used to use "asmtp" for authenticated SMTP, without any
5667 indication (in the protocol name) for TLS use. Now it follows the RFC and
5668 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5669 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5670 names appear in log lines as well as in Received: header lines.
5672 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5673 extensions, and comparison tests, and to fix some bugs.
5675 14. Changes to the "personal" filter test:
5677 (1) The test was buggy in that it was just doing the equivalent of
5678 "contains" tests on header lines. For example, if a user's address was
5679 anne@some.where, the "personal" test would incorrectly be true for
5681 To: susanne@some.where
5683 This test is now done by extracting each address from the header in turn,
5684 and checking the entire address. Other tests that are part of "personal"
5685 are now done using regular expressions (for example, to check local parts
5686 of addresses in From: header lines).
5688 (2) The list of non-personal local parts in From: addresses has been
5689 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5690 taken from the Sieve specification recommendations.
5692 (3) If the message contains any header line starting with "List-" it is
5693 treated as non-personal.
5695 (4) The test for "circular" in the Subject: header line has been removed
5696 because it now seems ill-conceived.
5698 15. Minor typos in src/EDITME comments corrected.
5700 16. Installed latest exipick from John Jetmore.
5702 17. If headers_add on a router specified a text string that was too long for
5703 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5704 of string_sprintf() is now avoided.
5706 18. $message_body_size was not set (it was always zero) when running the DATA
5707 ACL and the local_scan() function.
5709 19. For the "mail" command in an Exim filter, no default was being set for
5710 the once_repeat time, causing a random time value to be used if "once" was
5711 specified. (If the value happened to be <= 0, no repeat happened.) The
5712 default is now 0s, meaning "never repeat". The "vacation" command was OK
5713 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5714 (I found it when inspecting the code).
5716 20. There is now an overall timeout for performing a callout verification. It
5717 defaults to 4 times the callout timeout, which applies to individual SMTP
5718 commands during the callout. The overall timeout applies when there is more
5719 than one host that can be tried. The timeout is checked before trying the
5720 next host. This prevents very long delays if there are a large number of
5721 hosts and all are timing out (e.g. when the network connections are timing
5722 out). The value of the overall timeout can be changed by specifying an
5723 additional sub-option for "callout", called "maxwait". For example:
5725 verify = sender/callout=5s,maxwait=20s
5727 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5728 to the end before writing, but this should make it even safer).
5730 22. Exim was forgetting that it had advertised PIPELINING for the second and
5731 subsequent messages on an SMTP connection. It was also not resetting its
5732 memory on STARTTLS and an internal HELO.
5734 23. When Exim logs an SMTP synchronization error within a session, it now
5735 records whether PIPELINING has been advertised or not.
5737 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5738 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5739 rather than long int.
5741 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5743 26. Added the never_mail option to autoreply.
5749 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5750 crash if the getsockname() call failed; this can happen if a connection is
5751 closed very soon after it is established. The problem was simply in the
5752 order in which certain operations were done, causing Exim to try to write
5753 to the SMTP stream before it had set up the file descriptor. The bug has
5754 been fixed by making things happen in the correct order.
5760 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5761 before the connection was closed, thus losing the rejection response.
5763 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5764 some early Solaris releases, but causes trouble in current releases where
5765 socklen_t is defined.
5767 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5770 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5773 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5774 that releasing the top of it at the end releases what was used for sub-
5775 expansions (unless the block got too big). However, discard this block if
5776 the first thing is a variable or header, so that we can use its block when
5777 it is dynamic (useful for very large $message_headers, for example).
5779 6. Lookups now cache *every* query, not just the most recent. A new, separate
5780 store pool is used for this. It can be recovered when all lookup caches are
5781 flushed. Lookups now release memory at the end of their result strings.
5782 This has involved some general refactoring of the lookup sources.
5784 7. Some code has been added to the store_xxx() functions to reduce the amount
5785 of flapping under certain conditions.
5787 8. log_incoming_interface used to affect only the <= reception log lines. Now
5788 it causes the local interface and port to be added to several more SMTP log
5789 lines, for example "SMTP connection from", and rejection lines.
5791 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5793 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5795 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5796 could be overwritten at the end of the current message (or the start of a
5797 new message if it was set in a HELO ACL). The value is now preserved for
5798 the duration of the SMTP connection.
5800 12. If a transport had a headers_rewrite setting, and a matching header line
5801 contained an unqualified address, that address was qualified, even if it
5802 did not match any rewriting rules. The underlying bug was that the values
5803 of the flags that permit the existence of unqualified sender and recipient
5804 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5805 non-local messages, and by -bnq for local messages) were not being
5806 preserved with the message after it was received.
5808 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5809 "next input=" as part of the text comprising the host identity instead of
5810 the correct text. The code was using the same buffer for two different
5811 strings. However, depending on which order the printing function evaluated
5812 its arguments, the bug did not always show up. Under Linux, for example, my
5813 test suite worked just fine.
5815 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5816 doesn't work with some older versions of Perl. It has been changed to "my",
5817 which in any case is probably the better facility to use.
5819 15. A really picky compiler found some instances of statements for creating
5820 error messages that either had too many or two few arguments for the format
5823 16. The size of the buffer for calls to the DNS resolver has been increased
5824 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5825 for addresses that have a lot of PTR records. This alleviates a problem; it
5826 does not fully solve it.
5828 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5829 buffer now truncates the list and logs the incident, which is the same
5830 action as happens when Exim is looking up a host name and its aliases.
5831 Previously in this situation something unpredictable would happen;
5832 sometimes it was "internal error: store_reset failed".
5834 18. If a server dropped the connection unexpectedly when an Exim client was
5835 using GnuTLS and trying to read a response, the client delivery process
5836 crashed while trying to generate an error log message.
5838 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5839 string, for example:
5841 warn message = H1: something\nH2: something
5843 the text was added as a single header line from Exim's point of view
5844 though it ended up OK in the delivered message. However, searching for the
5845 second and subsequent header lines using $h_h2: did not work. This has been
5846 fixed. Similarly, if a system filter added multiple headers in this way,
5847 the routers could not see them.
5849 20. Expanded the error message when iplsearch is called with an invalid key to
5850 suggest using net-iplsearch in a host list.
5852 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5853 ACLs are no longer actually imposed (and a message to that effect is
5856 22. If a "gecos" field in a passwd entry contained escaped characters, in
5857 particular, if it contained a \" sequence, Exim got it wrong when building
5858 a From: or a Sender: header from that name. A second bug also caused
5859 incorrect handling when an unquoted " was present following a character
5860 that needed quoting.
5862 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5863 was not being matched caselessly.
5865 24. Arranged for all hyphens in the exim.8 source to be escaped with
5868 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5869 properly. Recipient callout cache records were still being keyed to include
5870 the sender, even when use_sender was set false. This led to far more
5871 callouts that were necessary. The sender is no longer included in the key
5872 when use_sender is false.
5874 26. Added "control = submission" modifier to ACLs.
5876 27. Added the ${base62d: operator to decode base 62 numbers.
5878 28. dnsdb lookups can now access SRV records.
5880 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5881 the configuration file.
5883 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5884 (-v) mode. This makes the output for a verbose queue run more intelligible.
5886 31. Added a use_postmaster feature to recipient callouts.
5888 32. Added the $body_zerocount variable, containing the number of binary zero
5889 bytes in the message body.
5891 33. The time of last modification of the "new" subdirectory is now used as the
5892 "mailbox time last read" when there is a quota error for a maildir
5895 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5897 35. Added +ignore_unknown as a special item in host lists.
5899 36. Code for decoding IPv6 addresses in host lists is now included, even if
5900 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5901 address was recognized as an IP address, but was then not correctly decoded
5902 into binary, causing unexpected and incorrect effects when compared with
5909 1. Very minor rewording of debugging text in manualroute to say "list of
5910 hosts" instead of "hostlist".
5912 2. If verify=header_syntax was set, and a header line with an unqualified
5913 address (no domain) and a large number of spaces between the end of the
5914 name and the colon was received, the reception process suffered a buffer
5915 overflow, and (when I tested it) crashed. This was caused by some obsolete
5916 code that should have been removed. The fix is to remove it!
5918 3. When running in the test harness, delay a bit after writing a bounce
5919 message to get a bit more predictability in the log output.
5921 4. Added a call to search_tidyup() just before forking a reception process. In
5922 theory, someone could use a lookup in the expansion of smtp_accept_max_
5923 per_host which, without the tidyup, could leave open a database connection.
5925 5. Added the variables $recipient_data and $sender_data which get set from a
5926 lookup success in an ACL "recipients" or "senders" condition, or a router
5927 "senders" option, similar to $domain_data and $local_part_data.
5929 6. Moved the writing of debug_print from before to after the "senders" test
5932 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5933 problems for message scanning, either using a data ACL, or using
5934 local_scan() because the Received: header was not generated till after they
5935 were called (in order to set the time as the time of reception completion).
5936 I have revised the way this works. The header is now generated after the
5937 body is received, but before the ACL or local_scan() are called. After they
5938 are run, the timestamp in the header is updated.
5944 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5945 before starting a queue runner without re-exec. This happened only when
5946 deliver_drop_privilege was set or when the Exim user was set to root. The
5947 effect of the bug was that timeouts during subsequent deliveries caused
5948 crashes instead of being properly handled. The handler is now left at its
5949 default (and expected) setting.
5951 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5952 message, again when deliver_drop_privilege is set or Exim is run as root.
5953 The bug described in (1) was not present in this case, but the tidying up
5954 of the other signals was missing. I have made the two cases consistent.
5956 3. The ignore_target_hosts setting on a manualroute router was being ignored
5957 for hosts that were looked up using the /MX notation.
5959 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5962 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5963 operated on the sender address. After changing the $sender_address to <>
5964 for the sender address verify, Exim was re-instated it as the original
5965 (before rewriting) address, but remembering that it had rewritten it, so it
5966 wasn't rewriting it again. This bug also had the effect of breaking the
5967 sender address verification caching when the sender address was rewritten.
5969 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5970 This has been changed so that if the ip literal address matches
5971 ignore_target_hosts, the router declines.
5973 7. Added expansion conditions match_domain, match_address, and match_local_
5974 part (NOT match_host).
5976 8. The placeholder for the Received: header didn't have a length field set.
5978 9. Added code to Exim itself and to exim_lock to test for a specific race
5979 condition that could lead to file corruption when using MBX delivery. The
5980 issue is with the lockfile that is created in /tmp. If this file is removed
5981 after a process has opened it but before that process has acquired a lock,
5982 there is the potential for a second process to recreate the file and also
5983 acquire a lock. This could lead to two Exim processes writing to the file
5984 at the same time. The added code performs the same test as UW imapd; it
5985 checks after acquiring the lock that its file descriptor still refers to
5986 the same named file.
5988 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5989 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5992 11. Added the smtp_active_hostname option. If used, this will typically be made
5993 to depend on the incoming interface address. Because $interface_address is
5994 not set up until the daemon has forked a reception process, error responses
5995 that can happen earlier (such as "too many connections") no longer contain
5998 12. If an expansion in a condition on a "warn" statement fails because a lookup
5999 defers, the "warn" statement is abandoned, and the next ACL statement is
6000 processed. Previously this caused the whole ACL to be aborted.
6002 13. Added the iplsearch lookup type.
6004 14. Added ident_timeout as a log selector.
6006 15. Added tls_certificate_verified as a log selector.
6008 16. Added a global option tls_require_ciphers (compare the smtp transport
6009 option of the same name). This controls incoming TLS connections.
6011 17. I finally figured out how to make tls_require_ciphers do a similar thing
6012 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6013 before starting the TLS session.
6015 18. Tabs are now shown as \t in -bP output.
6017 19. If the log selector return_path_on_delivery was set, Exim crashed when
6018 bouncing a message because it had too many Received: header lines.
6020 20. If two routers both had headers_remove settings, and the first one included
6021 a superfluous trailing colon, the final name in the first list and the
6022 first name in the second list were incorrectly joined into one item (with a
6023 colon in the middle).
6029 1. Added -C and -D options to the exinext utility, mainly to make it easier
6030 to include in the automated testing, but these could be helpful when
6031 multiple configurations are in use.
6033 2. The exinext utility was not formatting the output nicely when there was
6034 an alternate port involved in the retry record key, nor when there was a
6035 message id as well (for retries that were specific to a specific message
6036 and a specific host). It was also confused by IPv6 addresses, because of
6037 the additional colons they contain. I have fixed the IPv4 problem, and
6038 patched it up to do a reasonable job for IPv6.
6040 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6041 delivery, the log line now contains "pipelined" if PIPELINING was used.
6043 4. An SMTP transport process used to panic and die if the bind() call to set
6044 an explicit outgoing interface failed. This has been changed; it is now
6045 treated in the same way as a connect() failure.
6047 5. A reference to $sender_host_name in the part of a conditional expansion
6048 that was being skipped was still causing a DNS lookup. This no longer
6051 6. The def: expansion condition was not recognizing references to header lines
6052 that used bh_ and bheader_.
6054 7. Added the _cache feature to named lists.
6056 8. The code for checking quota_filecount in the appendfile transport was
6057 allowing one more file than it should have been.
6059 9. For compatibility with Sendmail, the command line option
6067 and sets the incoming protocol and host name (for trusted callers). The
6068 host name and its colon can be omitted when only the protocol is to be set.
6069 Note the Exim already has two private options, -pd and -ps, that refer to
6070 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6071 "s", but I don't think that's a major issue.
6073 10. A number of refactoring changes to the code, none of which should affect
6076 (a) The number of logging options was getting close to filling up the
6077 32-bit word that was used as a bit map. I have split them into two classes:
6078 those that are passed in the argument to log_write(), and those that are
6079 only ever tested independently outside of that function. These are now in
6080 separate 32-bit words, so there is plenty of room for expansion again.
6081 There is no change in the user interface or the logging behaviour.
6083 (b) When building, for example, log lines, the code previously used a
6084 macro that called string_cat() twice, in order to add two strings. This is
6085 not really sufficiently general. Furthermore, there was one instance where
6086 it was actually wrong because one of the argument was used twice, and in
6087 one call a function was used. (As it happened, calling the function twice
6088 did not affect the overall behaviour.) The macro has been replaced by a
6089 function that can join an arbitrary number of extra strings onto a growing
6092 (c) The code for expansion conditions now uses a table and a binary chop
6093 instead of a serial search (which was left over from when there were very
6094 few conditions). Also, it now recognizes conditions like "pam" even when
6095 the relevant support is not compiled in: a suitably worded error message is
6096 given if an attempt is made to use such a condition.
6098 11. Added ${time_interval:xxxxx}.
6100 12. A bug was causing one of the ddress fields not to be passed back correctly
6101 from remote delivery subprocesses. The field in question was not being
6102 subsequently used, so this caused to problems in practice.
6104 13. Added new log selectors queue_time and deliver_time.
6106 14. Might have fixed a bug in maildirsizefile handling that threw up
6107 "unexpected character" debug warnings, and recalculated the data
6108 unnecessarily. In any case, I expanded the warning message to give more
6111 15. Added the message "Restricted characters in address" to the statements in
6112 the default ACL that block characters like @ and % in local parts.
6114 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6115 Three changes have been made:
6117 (a) There was a serious bug; a negative response to MAIL caused the whole
6118 recipient domain to be cached as invalid, thereby blocking all messages
6119 to all local parts at the same domain, from all senders. This bug has
6120 been fixed. The domain is no longer cached after a negative response to
6121 MAIL if the sender used is not empty.
6123 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6126 (c) A new callout option, "use_sender" has been added for people who want
6127 the modified behaviour.
6133 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6136 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6137 indeed breaks things for older releases.
6139 3. Added additional logging to the case where there is a problem reading data
6140 from a filter that is running in a subprocess using a pipe, in order to
6141 try to track down a specific problem.
6143 4. Testing facility fudge: when running in the test harness and attempting
6144 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6145 getting "No route to host". Convert this to a timeout.
6147 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6150 6. Some OS don't have socklen_t but use size_t instead. This affects the
6151 fifth argument of getsockopt() amongst other things. This is now
6152 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6153 can be set for individual OS. I have set it for SunOS5, OSF1, and
6154 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6155 some earlier ones do not.
6157 7. Change 4.30/15 was not doing the test caselessly.
6159 8. The standard form for an IPv6 address literal was being rejected by address
6160 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6161 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6162 this, as well as the form without the "IPv6" on the front (but only when
6163 address literals are enabled, of course).
6165 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6167 10. Exim crashed if a message with an empty sender address specified by -f
6168 encountered a router with an errors_to setting. This could be provoked only
6169 by a command such as
6173 where an empty string was supplied; "<>" did not hit this bug.
6175 11. Installed PCRE release 4.5.
6177 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6178 remained set. It is now erased.
6180 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6181 times from message ids (which are base 36 rather than the normal 62).
6183 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6184 were being counted as actual protocol errors, and logged if the log
6185 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6186 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6187 and DATA following a set of rejected RCPTs do not count as protocol errors.
6188 In other words, Exim assumes they were pipelined, though this may not
6189 actually be the case. Of course, in all cases the client gets an
6190 appropriate error code.
6192 15. If a lookup fails in an ACL condition, a message about the failure may
6193 be available; it is used if testing the ACL cannot continue, because most
6194 such messages specify what the cause of the deferral is. However, some
6195 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6196 that caused an old message to be retained and used if a later statement
6197 caused a defer, replacing the real cause of the deferral.
6199 16. If an IP address had so many PTR records that the DNS lookup buffer
6200 was not large enough to hold them, Exim could crash while trying to process
6201 the truncated data. It now detects and logs this case.
6203 17. Further to 4.21/58, another change has been made: if (and only if) the
6204 first line of a message (the first header line) ends with CRLF, a bare LF
6205 in a subsequent header line has a space inserted after it, so as not to
6206 terminate the header.
6208 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6209 unnecessarily, used atoi() instead of strtol(), and didn't check the
6210 termination when getting file sizes from file names by regex.
6212 19. Completely re-implemented the support for maildirsize files, in the light
6213 of a number of problems with the previous contributed implementation
6214 (4.30/29). In particular:
6216 . If the quota is zero, the maildirsize file is maintained, but no quota is
6219 . If the maildir directory does not exist, it is created before any attempt
6220 to write a maildirsize file.
6222 . The quota value in the file is just a cache; if the quota is changed in
6223 the transport, the new value overrides.
6225 . A regular expression is available for excluding directories from the
6228 20. The autoreply transport checks the characters in options that define the
6229 message's headers; it allows continued headers, but it was checking with
6230 isspace() after an embedded newline instead of explicitly looking for a
6233 21. If all the "regular" hosts to which an address was routed had passed their
6234 expiry times, and had not reached their retry times, the address was
6235 bounced, even if fallback hosts were defined. Now Exim should go on to try
6238 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6239 equivalent code in the SMTP transport. Some hosts send humungous responses
6240 to HELO/EHLO, more than 1024 it seems.
6242 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6243 gives compiler warnings in some environments. I've now done it "properly",
6246 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6247 (because of problems with the built-in one) was declared to return uschar *
6248 instead of char *, causing compiler failure.
6250 25. Fixed a file descriptor leak when processing alias/forward files.
6252 26. Fixed a minor format string issue in dbfn.c.
6254 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6256 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6257 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6260 29. When a sender address is verified, it is cached, to save repeating the test
6261 when there is more than one recipient in a message. However, when the
6262 verification involves a callout, it is possible for different callout
6263 options to be set for different recipients. It is too complicated to keep
6264 track of this in the cache, so now Exim always runs a verification when a
6265 callout is required, relying on the callout cache for the optimization.
6266 The overhead is duplication of the address routing, but this should not be
6269 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6270 to be verified with callout=postmaster, and the main callout worked but the
6271 postmaster check failed, the verification correctly failed. However, if a
6272 subsequent RCPT command asked for sender verification *without* the
6273 postmaster check, incorrect caching caused this verification also to fail,
6276 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6277 it was not caching the DNS options (qualify_single, search_parents) that
6278 were used when the lookup failed. A subsequent lookup with different
6279 options therefore always gave the same answer, though there were cases
6280 where it should not have. (Example: a "domains = !$mx_any" option on a
6281 dnslookup router: the "domains" option is always processed without any
6282 widening, but the router might have qualify_single set.) Now Exim uses the
6283 cached value only when the same options are set.
6285 32. Added John Jetmore's "exipick" utility to the distribution.
6287 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6288 than a timeout (e.g. a certificate is required, and is not provided), an
6289 Exim server now closes the connection immediately. Previously it waited for
6290 the client to close - but if the client is SSL, it seems that they each
6291 wait for each other, leading to a delay before one of them times out.
6293 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6294 maintained 0.8.x compatibility because I don't think many are using it, and
6295 it is clearly obsolete.
6297 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6300 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6301 client certificate was expired. A simple patch fixes this, though I don't
6302 understand the full logic of why the verify callback is called multiple
6305 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6306 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6307 which causes problems with some clients (such as the Certicom SSL Plus
6308 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6309 disables the coutermeasure allowing Eudora to connect."
6311 38. Exim was not checking that a write() to a log file succeeded. This could
6312 lead to Bad Things if a log got too big, in particular if it hit a file
6313 size limit. Exim now panics and dies if it cannot write to a log file, just
6314 as it does if it cannot open a log file.
6316 39. Modified OS/Makefile-Linux so that it now contains
6318 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6320 The two -D definitions ensure that Exim is compiled with large file
6321 support, which makes it possible to handle log files that are bigger than
6324 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6325 instance) a domain was checked against a named list that involved a lookup,
6326 causing $domain_data to be set, then another domain was checked against the
6327 same list, then the first domain was re-checked, the value of $domain_data
6328 after the final check could be wrong. In particular, if the second check
6329 failed, it could be set empty. This bug probably also applied to
6332 41. The strip_trailing_dot option was not being applied to the address given
6333 with the -f command-line option.
6335 42. The code for reading a message's header from the spool was incrementing
6336 $received_count, but never initializing it. This meant that the value was
6337 incorrect (doubled) while delivering a message in the same process in which
6338 it was received. In the most common configuration of Exim, this never
6339 happens - a fresh exec is done - but it can happen when
6340 deliver_drop_privilege is set.
6342 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6343 it now includes up to 150 characters of the unexpected data in the log
6346 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6347 and building data strings. The size of both of these buffers was 10 000
6348 bytes - far larger than anybody would *ever* want, thought I. Needless to
6349 say, somebody hit the limit. I have increased the maximum line length to
6350 20 000 and the maximum data length of concatenated lines to 100 000. I have
6351 also fixed two bugs, because there was no checking on these buffers. Tsk,
6352 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6353 buffer is too small.
6355 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6356 lsearch lookups. Now it does.
6358 46. When parsing a route_list item in a manualroute router, a fixed-length
6359 buffer was used for the list of hosts. I made this 1024 bytes long,
6360 thinking that nobody would ever have a list of hosts that long. Wrong.
6361 Somebody had a whole pile of complicated expansion conditions, and the
6362 string was silently truncated, leading to an expansion error. It turns out
6363 that it is easier to change to an unlimited length (owing to other changes
6364 that have happened since this code was originally written) than to build
6365 structure for giving a limitation error. The length of the item that
6366 expands into the list of hosts is now unlimited.
6368 47. The lsearch lookup could not handle data where the length of text line was
6369 more than 4095 characters. Such lines were truncated, leading to shortened
6370 data being returned. It should now handle lines of any length.
6372 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6373 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6376 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6377 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6378 no longer happens in comment lines. A list of replacements is now placed
6379 at the head of all of the source files, except those whose only change is
6380 to replace PERL_COMMAND in the very first #! line.
6382 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6383 messages on the queue, with a bottom-up merge sort, using code contributed
6384 by Michael Haardt. This should make operations like -bp somewhat faster on
6385 large queues. It won't affect queue runners, except when queue_run_in_order
6388 51. Installed eximstats 1.31 in the distribution.
6390 52. Added support for SRV lookups to the dnslookup router.
6392 53. If an ACL referred to $message_body or $message_body_end, the value was not
6393 reset for any messages that followed in the same SMTP session.
6395 54. The store-handling optimization for building very long strings was not
6396 differentiating between the different store pools. I don't think this
6397 actually made any difference in practice, but I've tidied it.
6399 55. While running the routers to verify a sender address, $sender_address
6400 was still set to the sender address. This is wrong, because when routing to
6401 send a bounce to the sender, it would be empty. Therefore, I have changed
6402 it so that, while verifying a sender address, $sender_address is set to <>.
6403 (There is no change to what happens when verifying a recipient address.)
6405 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6406 target A or AAAA records (if not already returned) without resetting the
6407 qualify_single or search_parents options of the DNS resolver. These are
6408 inappropriate in this case because the targets of MX and SRV records must
6409 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6410 target that, when qualified, matched something in the local domain. These
6411 two options are now turned off when doing these lookups.
6413 57. It seems that at least some releases of Reiserfs (which does not have the
6414 concept of a fixed number of inodes) returns zero and not -1 for the
6415 number of available inodes. This interacted badly with check_spool_inodes,
6416 which assumed that -1 was the "no such thing" setting. What I have done is
6417 to check that the total number of inodes is greater than zero before doing
6418 the test of how many are available.
6420 58. When a "warn" ACL statement has a log_message modifier, the message is
6421 remembered, and not repeated. This is to avoid a lot of repetition when a
6422 message has many recipients that cause the same warning to be written.
6423 However, Exim was preserving the list of already written lines for an
6424 entire SMTP session, which doesn't seem right. The memory is now reset if a
6425 new message is started.
6427 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6428 debugging output unless log_rewrite was also set.
6430 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6431 of (int)(handle) when we know that handle contains (void *)(-1).
6433 61. The Exim daemon panic-logs an error return when it closes the incoming
6434 connection. However "connection reset by peer" seems to be common, and
6435 isn't really an error worthy of noting specially, so that particular error
6438 62. When Exim is trying to find all the local interfaces, it used to panic and
6439 die if the ioctl to get the interface flags failed. However, it seems that
6440 on at least one OS (Solaris 9) it is possible to have an interface that is
6441 included in the list of interfaces, but for which you get a failure error
6442 for this call. This happens when the interface is not "plumbed" into a
6443 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6444 failure of the "get flags" call assumes that the interface is down.
6446 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6447 makes life easier for people who are doing arithmetic on fields extracted
6448 from dates, where you often get leading zeros that should not be
6449 interpreted as octal.
6451 64. Added qualify_domain to the redirect router, to override the global
6454 65. If a pathologically long header line contained very many addresses (the
6455 report of this problem mentioned 10 000) and each of them was rewritten,
6456 Exim could use up a very large amount of memory. (It kept on making new
6457 copies of the header line as it rewrote, and never released the old ones.)
6458 At the expense of a bit more processing, the header rewriting function has
6459 been changed so that it no longer eats memory in this way.
6461 66. The generation of the Received: header has been moved from the time that a
6462 message starts to be received, to the time that it finishes. The timestamp
6463 in the Received: header should now be very close to that of the <= log
6464 line. There are two side-effects of this change:
6466 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6467 logged header lines no longer include the local Received: line, because
6468 it has not yet been created. The same applies to a copy of the message
6469 that is returned to a non-SMTP sender when a message is rejected.
6471 (b) When a filter file is tested using -bf, no additional Received: header
6472 is added to the test message. After some thought, I decided that this
6475 This change does not affect the value of $received_for. It is still set
6476 after address rewriting, but before local_scan() is called.
6478 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6480 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6481 gave an unhelpful panic error message, and a defer error. I have managed to
6482 change this behaviour so that it now rejects any supplied certificate,
6483 which seems right, as the list of acceptable certificates is empty.
6485 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6486 gave an unhelpful defer error. I have not managed to make this reject any
6487 supplied certificates, but the error message it gives is "no certificate
6488 supplied", which is not helpful.
6490 70. exigrep's output now also includes lines that are not associated with any
6491 message, but which match the given pattern. Implemented by a patch from
6492 Martin Sluka, which also tidied up the Perl a bit.
6494 71. Recipient callout verification, like sender verification, was using <> in
6495 the MAIL FROM command. This isn't really the right thing, since the actual
6496 sender may affect whether the remote host accepts the recipient or not. I
6497 have changed it to use the actual sender in the callout; this means that
6498 the cache record is now keyed on a recipient/sender pair, not just the
6499 recipient address. There doesn't seem to be a real danger of callout loops,
6500 since a callout by the remote host to check the sender would use <>.
6501 [SEE ABOVE: changed after hitting problems.]
6503 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6504 temporary errors. However, in the case of such a code being given after
6505 the end of a data transmission (i.e. after ".") Exim was failing to write
6506 a retry record for the message. (Yes, there was some broken host that was
6507 actually sending 8xx at this point.)
6509 73. An unknown lookup type in a host list could cause Exim to panic-die when
6510 the list was checked. (An example that provoked this was putting <; in the
6511 middle of a list instead of at the start.) If this happened during a DATA
6512 ACL check, a -D file could be left lying around. This kind of configuration
6513 error no longer causes Exim to die; instead it causes a defer error. The
6514 incident is still logged to the main and panic logs.
6516 74. Buglet left over from Exim 3 conversion. The message "too many messages
6517 in one connection" was written to the rejectlog but not the mainlog, except
6518 when address rewriting (yes!) was being logged.
6520 75. Added write_rejectlog option.
6522 76. When a system filter was run not as root (that is, when system_filter_user
6523 was set), the values of the $n variables were not being returned to the
6524 main process; thus, they were not subsequently available in the $sn
6527 77. Added +return_path_on_delivery log selector.
6529 78. A connection timeout was being treated differently from recipients deferred
6530 when testing hosts_max_try with a message that was older than the host's
6531 retry timeout. (The host should not be counted, thus allowing all hosts to
6532 be tried at least once before bouncing.) This may have been the cause of an
6533 occasionally reported bug whereby a message would remain on the queue
6534 longer than the retry timeout, but would be bounced if a delivery was
6535 forced. I say "may" because I never totally pinned down the problem;
6536 setting up timeout/retry tests is difficult. See also the next item.
6538 79. The ultimate address timeout was not being applied to errors that involved
6539 a combination of host plus message (for example, a timeout on a MAIL
6540 command). When an address resolved to a number of possible hosts, and they
6541 were not all tried for each delivery (e.g. because of hosts_max_try), a
6542 message could remain on the queue longer than the retry timeout.
6544 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6547 81. Fixed an obscure SMTP outgoing bug which required at least the following
6548 conditions: (a) there was another message waiting for the same server;
6549 (b) the server returned 5xx to all RCPT commands in the first message so
6550 that the message was not completed; (c) the server dropped the connection
6551 or gave a negative response to the RSET that Exim sends to abort the
6552 transaction. The observed case was a dropped connection after DATA that had
6553 been sent in pipelining mode. That is, the server had advertised PIPELINING
6554 but was not implementing it correctly. The effect of the bug was incorrect
6555 behaviour, such as trying another host, and this could lead to a crash.
6561 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6562 and daemon.c were passed as pointers to ints; they should have been
6563 pointers to socklen_t variables (which are typically unsigned ints).
6565 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6568 3. Fixed a really odd bug that affected only the testing scheme; patching a
6569 certain fixed string in the binary changed the value of another string that
6570 happened to be identical to the end of the original first string.
6572 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6573 name", it returns that address as the IP address. On some operating
6574 systems (e.g. Solaris), it also passes back the IP address string as the
6575 "host name". However, on others (e.g. Linux), it passes back an empty
6576 string. Exim wasn't checking for this, and was changing the host name to an
6577 empty string, assuming it had been canonicalized.
6579 5. Although rare, it is permitted to have more than one PTR record for a given
6580 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6581 all the names associated with an address, because they do in Solaris.
6582 However, it seems that they do not in Linux for data that comes from the
6583 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6584 I found this out when I moved to a new Linux workstation and tried to run
6585 the Exim test suite.
6587 To get round this problem I have changed the code so that it now does its
6588 own call to the DNS to look up PTR records when searching for a host name.
6589 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6590 addresses that are only in /etc/hosts are still found.
6592 This behaviour is, however, controlled by an option called host_lookup_
6593 order, which defaults to "bydns:byaddr". If people want to use the other
6594 order, or indeed, just use one or the other means of lookup, they can
6595 specify it in this variable.
6597 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6598 some operating systems, this comes back from gethostbyaddr() as an empty
6599 string, and this is what Exim used to test for. However, it seems that in
6600 other systems, "." is yielded. Exim now tests for this case too.
6602 7. The values of check_spool_space and check_log_space are now held internally
6603 as a number of kilobytes instead of an absolute number of bytes. If a
6604 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6605 kilobyte. This means that much larger values can be stored.
6607 8. Exim monitor: an attempt to get the action menu when not actually pointing
6608 at a message produces an empty menu entitled "No message selected". This
6609 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6610 no entries in it ("Shell widget menu has zero width and/or height"). So I
6611 have added a single, blank menu entry in this case.
6613 9. Added ${quote_local_part.
6615 10. MIME decoding is now applied to the contents of Subject: header lines when
6618 11. Now that a reference to $sender_host_address automatically causes a reverse
6619 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6620 host lookup before query-style lookups in lists that might use this
6621 variable. This has therefore been abolished, and the "net-" prefix is no
6622 longer necessary for query-style lookups.
6624 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6625 have been a typo for LFLAGS, so it has been changed.
6627 13. The install script calls Exim with "-C /dev/null" in order to find the
6628 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6629 to be output. However, since Exim outputs its version number before the
6630 error, it didn't break the script. It just looked ugly. I fixed this by
6631 always allowing "-C /dev/null" if the caller is root.
6633 14. Ignore overlarge ACL variable number when reading spool file - insurance
6634 against a later release with more variables having written the file.
6636 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6637 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6638 this, as well as the form without the "IPv6" on the front.
6640 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6641 OS/Makefile-Darwin file.
6643 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6644 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6646 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6647 variable. However, some OS use ints for inodes. I've added cast to long int
6648 to get rid of the compiler warning.
6650 19. I had forgotten to lock out "/../" in configuration file names when
6651 ALT_CONFIG_PREFIX was set.
6653 20. Routers used for verification do not need to specify transports. However,
6654 if such a router generated a host list, and callout was configured, Exim
6655 crashed, because it could not find a port number from the (non-existent)
6656 transport. It now assumes port 25 in this circumstance.
6658 21. Added the -t option to exigrep.
6660 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6661 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6662 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6663 them is set, LOOKUP_LSEARCH is forced.
6665 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6667 24. Added sender and host information to the "rejected by local_scan()" log
6668 line; previously there was no indication of these.
6670 25. Added .include_if_exists.
6672 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6673 Linux. It turns out that not all file systems support this. Apparently some
6674 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6675 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6676 not supported on the file descriptor, is now ignored when Exim is trying to
6677 sync a directory. This applies only to Linux.
6679 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6681 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6682 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6685 29. Added support for maildirsize files from supplied patch (modified a bit).
6687 30. The use of :fail: followed by an empty string could lead Exim to respond to
6688 sender verification failures with (e.g.):
6690 550 Verification failed for <xxx>
6691 550 Sender verify failed
6693 where the first response line was missing the '-' that indicates it is not
6694 the final line of the response.
6696 31. The loop for finding the name of the user that called Exim had a hardwired
6697 limit of 10; it now uses the value of finduser_retries, which is used for
6698 all other user lookups.
6700 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6703 33. Exim was neglecting to zero errno before one call of strtol() when
6704 expanding a string and expecting an integer value. On some systems this
6705 resulted in spurious "integer overflow" errors. Also, it was casting the
6706 result into an int without checking.
6708 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6709 did not work. The code looks as if it has *never* worked, though it appears
6710 to have been documented since at least release 1.62. I have made it work.
6712 35. The "timeout_DNS" error in retry rules, also documented since at least
6713 1.62, also never worked. As it isn't clear exactly what this means, and
6714 clearly it isn't a major issue, I have abolished the feature by treating it
6715 as "timeout", and writing a warning to the main and panic logs.
6717 36. The display of retry rules for -brt wasn't always showing the error code
6720 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6721 timeout_connect_A, timeout_connect_MX.
6723 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6724 to the empty sender.
6726 39. The daemon was not analysing the content of -oX till after it had closed
6727 stderr and disconnected from the controlling terminal. This meant that any
6728 syntax errors were only noted on the panic log, and the return code from
6729 the command was 0. By re-arranging the code a little, I've made the
6730 decoding happen first, so such errors now appear on stderr, and the return
6731 code is 1. However, the actual setting up of the sockets still happens in
6732 the disconnected process, so errors there are still only recorded on the
6735 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6736 connections (as happens on some IP stacks) was logged at start up time as
6737 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6738 it from "IPv6 and IPv4", which means that two separate sockets are being
6741 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6742 says whether AF_INET or AF_INET6 was passed as an argument.
6744 42. Exiwhat output was messed up when time zones were included in log
6747 43. Exiwhat now gives more information about the daemon's listening ports,
6748 and whether -tls-on-connect was used.
6750 44. The "port" option of the smtp transport is now expanded.
6752 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6753 silently ignored. Now an error message is written to the main and panic
6756 46. There's a new ACL modifier called "logwrite" which writes to a log file
6757 as soon as it is encountered.
6759 47. Added $local_user_uid and $local_user_gid at routing time.
6761 48. Exim crashed when trying to verify a sender address that was being
6764 49. Exim was recognizing only a space character after ".include". It now also
6765 recognizes a tab character.
6767 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6768 extracting the relevant information from the specification. The man page no
6769 longer contains scrambled data for the -d option, and I've added a section
6770 at the front about calling Exim under different names.
6772 51. Added "extra_headers" argument to the "mail" command in filter files.
6774 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6777 53. Installed eximstats 1.29.
6779 54. Added transport_filter_timeout as a generic transport option.
6781 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6782 Cc: header lines. This was required by RFC 822, but it not required by RFC
6785 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6786 incoming messages that did not have them. Now it does so only if the
6787 message originates locally, that is, if there is no associated remote host
6788 address. When Resent- header lines are present, this applies to the Resent-
6789 lines rather than the non-Resent- lines.
6791 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6792 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6794 58. Messages for configuration errors now include the name of the main
6795 configuration file - useful now that there may be more than one file in a
6796 list (.included file names were always shown).
6798 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6799 for those rare installations that do not start the daemon as root or run it
6800 setuid root. I've cut out the call to initgroups() if the daemon is not
6803 60. The Exim user and group can now be bound into the binary as text strings
6804 that are looked up at the start of Exim's processing.
6806 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6808 62. Added $mailstore_basename variable.
6810 63. Installed patch to sieve.c from Michael Haardt.
6812 64. When Exim failed to open the panic log after failing to open the main log,
6813 the original message it was trying to log was written to stderr and debug
6814 output, but if they were not available (the usual case in production), it
6815 was lost. Now it is written to syslog before the two lines that record the
6816 failures to open the logs.
6818 65. Users' Exim filters run in subprocesses under the user's uid. It is
6819 possible for a "deliver" command or an alias in a "personal" command to
6820 provoke an address rewrite. If logging of address rewriting is configured,
6821 this fails because the process is not running as root or exim. There may be
6822 a better way of dealing with this, but for the moment (because 4.30 needs
6823 to be released), I have disabled address rewrite logging when running a
6824 filter in a non-root, non-exim process.
6830 1. The buildconfig auxiliary program wasn't quoting the value set for
6831 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6832 not defined. This bug was masked in 4.22 by the effect that was fixed in
6835 2. Some messages that were rejected after a message id was allocated were
6836 shown as "incomplete" by exigrep. It no longer does this for messages that
6837 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6839 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6840 have allow_domain_literals set, the ID did not get logged in the <= line.
6841 Domain literals are now always recognized in Message-ID: header lines.
6843 4. The first argument for a ${extract expansion item is the key name or field
6844 number. Leading and trailing spaces in this item were not being ignored,
6845 causing some misleading effects.
6847 5. When deliver_drop_privilege was set, single queue runner processes started
6848 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6849 same command in the process it spins off) were not dropping privilege.
6851 6. When the daemon running as "exim" started a queue runner, it always
6852 re-executed Exim in the spun-off process. This is a waste of effort when
6853 deliver_drop_privilege is set. The new process now just calls the
6854 queue-runner function directly.
6860 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6863 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6864 ignored. Though the use of -oP was forcing the writing of a pid file, it
6865 was always written to the default place.
6867 3. If the message "no IP address found for host xxxx" is generated during
6868 incoming verification, it is now followed by identification of the incoming
6869 connection (so you can more easily find what provoked it).
6871 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6873 5. Added some features to "harden" Exim a bit more against certain attacks:
6875 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6876 be put in Local/Makefile. This is like the never_users runtime option,
6877 but it cannot be overridden. The default setting is "root".
6879 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6880 prefix string with which any file named in a -C command line option
6883 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6884 is retained for -C and -D only if the caller of Exim is root. Without
6885 it, the exim user may also use -C and -D and retain privilege.
6887 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6888 command line option is disabled.
6890 6. Macro names set by the -D option must start with an upper case letter, just
6891 like macro names defined in the configuration file.
6893 7. Added "dereference=" facility to LDAP.
6895 8. Two instances of the typo "uknown" in the source files are fixed.
6897 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6898 the Configure-Makefile script screwed up while processing it.
6900 10. Incorporated PCRE 4.4.
6902 11. The SMTP synchronization check was not operating right at the start of an
6903 SMTP session. For example, it could not catch a HELO sent before the client
6904 waited for the greeting. There is now a check for outstanding input at the
6905 point when the greeting is written. Because of the duplex, asynchronous
6906 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6907 way, but not yet received, when the check is performed.
6909 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6910 on TCP/IP sockets, because this apparently causes some broken clients to
6913 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6914 unchanged) from the Cygwin maintainer.
6916 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6917 was supported instead of testing for mbx. Effectively a typo.
6919 15. The spa authenticator server code was not checking that the input it
6920 received was valid base64.
6922 16. The debug output line for the "set" modifier in ACLs was not showing the
6923 name of the variable that was being set.
6925 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6927 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6928 lookup on its own. Something else was needed to trigger the lookup. For
6929 example, a match in host_lookup or the need for a host name in a host list.
6930 Now, if $sender_host_name is referenced and the host name has not yet been
6931 looked up, a lookup is performed. If the lookup fails, the variable remains
6932 empty, and $host_lookup_failed is set to "1".
6934 19. Added "eqi" as a case-independent comparison operator.
6936 20. The saslauthd authentication condition could segfault if neither service
6937 nor realm was specified.
6939 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6940 error message that was logged was misleading, and incoming SMTP
6941 connections were dropped. The message is now more accurate, and temporary
6942 errors are given to SMTP connections.
6944 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6945 (except RSET) with a 421 error, until QUIT is received. However, it was
6946 failing to send a response to QUIT.
6948 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6949 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6950 ACL is now run after the helo_try_verify_hosts code.
6952 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6953 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6954 case-independent, so other case variants are also recognized). Apparently
6955 some systems use these upper case variants.
6957 25. If more than two messages were waiting for the same host, and a transport
6958 filter was specified for the transport, Exim sent two messages over the
6959 same TCP/IP connection, and then failed with "socket operation on non-
6960 socket" when it tried to send the third.
6962 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6964 27. The extern definition of crypt16() in expand.c was not being excluded when
6965 the OS had its own crypt16() function.
6967 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6968 as a preferred synonym for return_size_limit, both as an option and as an
6971 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6973 30. Changed the default configuration ACL to relax the local part checking rule
6974 for addresses that are not in any local domains. For these addresses,
6975 slashes and pipe symbols are allowed within local parts, but the sequence
6976 /../ is explicitly forbidden.
6978 31. SPA server authentication was not clearing the challenge buffer before
6981 32. log_message in a "warn" ACL statement was writing to the reject log as
6982 well as to the main log, which contradicts the documentation and doesn't
6983 seem right (because no rejection is happening). So I have stopped it.
6985 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6986 However, I am unable to do any testing of this.
6988 34. Fixed an infelicity in the appendfile transport. When checking directories
6989 for a mailbox, to see if any needed to be created, it was accidentally
6990 using path names with one or more superfluous leading slashes; tracing
6991 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6993 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6994 contents are added to the log line that is written for every discarded
6995 recipient. (Previously a log_message setting was ignored.)
6997 36. The ${quote: operator now quotes the string if it is empty.
6999 37. The install script runs exim in order to find its version number. If for
7000 some reason other than non-existence or emptiness, which it checks, it
7001 could not run './exim', it was installing it with an empty version number,
7002 i.e. as "exim-". This error state is now caught, and the installation is
7005 38. An argument was missing from the function that creates an error message
7006 when Exim fails to connect to the socket for saslauthd authentication.
7007 This could cause Exim to crash, or give a corrupted message.
7009 39. Added isip, isip4, and isip6 to ${if conditions.
7011 40. The ACL variables $acl_xx are now saved with the message, and can be
7012 accessed later in routers, transports, and filters.
7014 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7015 strings in the file are not string-expanded.
7017 42. If a MAIL command specified a SIZE value that was too large to fit into an
7018 int variable, the check against message_size_limit failed. Such values are
7019 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7020 day this will have to be increased, but I don't think I want to be around
7021 when emails are that large.
7028 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7029 iconv() is not standard in FreeBSD.
7031 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7032 IPv6 enabled. The observed symptom was a segmentation fault on return from
7033 the function os_common_find_running_interfaces() in src/os.c.
7035 3. In the check_special_case() function in daemon.c I had used "errno" as an
7036 argument name, which causes warnings on some systems. This was basically a
7037 typo, since it was named "eno" in the comments!
7039 4. The code that waits for the clock to tick (at a resolution of some fraction
7040 of a second) so as to ensure message-id uniqueness was always waiting for
7041 at least one whole tick, when it could have waited for less. [This is
7042 almost certainly not relevant at current processor speeds, where it is
7043 unlikely to ever wait at all. But we try to future-proof.]
7045 5. The function that sleeps for a time interval that includes fractions of a
7046 second contained a race. It did not block SIGALRM between setting the
7047 timer, and suspending (a couple of lines later). If the interval was short
7048 and the sigsuspend() was delayed until after it had expired, the suspension
7049 never ended. On busy systems this could lead to processes getting stuck for
7052 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7053 process, before it forks any delivery processes. The open lookup caching
7054 mechanism meant that the open file or database connection was passed into
7055 the delivery process. The problem was that delivery processes always tidy
7056 up cached lookup data. This could cause a problem for the next delivery
7057 process started by the queue runner, because the external queue runner
7058 process does not know about the closure. So the next delivery process
7059 still has data in the lookup cache. In the case of a file lookup, there was
7060 no problem because closing a file descriptor in a subprocess doesn't affect
7061 the parent. However, if the lookup was caching a connection to a database,
7062 the connection was closed, and the second delivery process was likely to
7063 see errors such as "PGSQL: query failed: server closed the connection
7064 unexpectedly". The problem has been fixed by closing all cached lookups
7065 in a queue runner before running a delivery process.
7067 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7068 seem to have the "const" qualifier which it has on other OS. I've
7071 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7072 *of the same type* (client or server) with the same public name that an
7073 error should be diagnosed.
7075 9. When Exim looked up a host name for an IP address, but failed to find the
7076 original IP address when looking up the host name (a safety check), it
7077 output the message "<ip address> does not match any IP for NULL", which was
7078 confusing, to say the least. The bug was that the host name should have
7079 appeared instead of "NULL".
7081 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7082 user that is built into the binary, and the -C or -D options is used, root
7083 privilege is dropped before the configuration file is read. In addition,
7084 logging is switched to stderr instead of the normal log files. If the
7085 configuration then re-defines the Exim user, the unprivileged environment
7086 is probably not what is expected, so Exim logs a panic warning message (but
7089 However, if deliver_drop_privilege is set, the unprivileged state may well
7090 be exactly what is intended, so the warning has been cut out in that case,
7091 and Exim is allowed to try to write to its normal log files.
7097 1. smtp_return_error_details was not giving details for temporary sender
7098 or receiver verification errors.
7100 2. Diagnose a configuration error if two authenticators have the same public
7103 3. Exim used not to create the message log file for a message until the first
7104 delivery attempt. This could be confusing when incoming messages were held
7105 for policy or load reasons. The message log file is now created at the time
7106 the message is received, and an initial "Received" line is written to it.
7108 4. The automatically generated man page for command line options had a minor
7109 bug that caused no ill effects; however, a more serious problem was that
7110 the procedure for building the man page automatically didn't always
7111 operate. Consequently, release 4.20 contains an out-of-date version. This
7112 shouldn't happen again.
7114 5. When building Exim with embedded Perl support, the script that builds the
7115 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7116 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7118 6. The freeze_tell option was not being used for messages that were frozen on
7119 arrival, either by an ACL or by local_scan().
7121 7. Added the smtp_incomplete_transaction log selector.
7123 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7124 was accepting AUTH without a new EHLO.
7126 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7127 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7130 10. Logging of TCP/IP connections (when configured) now happens in the main
7131 daemon process instead of the child process, so that the TCP/IP connection
7132 count is more accurate (but it can never be perfect).
7134 11. The use of "drop" in a nested ACL was not being handled correctly in the
7135 outer ACL. Now, if condition failure induced by the nested "drop" causes
7136 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7137 or "require"), the connection is dropped.
7139 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7140 that yield "discard" can now be used with an "accept" or a "discard" verb,
7141 but an error is generated for any others (because I can't see a useful way
7142 to define what should happen).
7144 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7145 are now processed in the same way as lines in the Exim configuration file.
7146 In particular, continuation lines are supported.
7148 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7150 15. Added -ti meaning -t -i.
7152 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7153 domains, and warn by logging if others are found.
7155 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7156 returned from GIFCONF when Exim is trying to find the list of interfaces on
7157 a host. The code in os.c has been modified to copy each ifreq to an aligned
7158 structure in all cases.
7160 Also, in some cases, the returned ifreq's were being copied to a 'struct
7161 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7162 means the last couple of bytes of an IPv6 address could be chopped if the
7163 ifreq contained only a normal sockaddr (14 bytes storage).
7165 18. Named domain lists were not supported in the hosts_treat_as_local option.
7166 An entry such as +xxxx was not recognized, and was treated as a literal
7169 19. Ensure that header lines added by a DATA ACL are included in the reject log
7170 if the ACL subsequently rejects the message.
7172 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7173 MD5 (which is deprecated).
7175 21. When testing a filter file using -bf, Exim was writing a message when it
7176 took the sender from a "From " line in the message, but it was not doing so
7177 when it took $return_path from a Return-Path: header line. It now does.
7179 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7180 with a valid header line field name (a series of printing characters
7181 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7183 23. Changed "disc" in the source to "disk" to conform to the documentation and
7184 the book and for uniformity.
7186 24. Ignore Sendmail's -Ooption=value command line item.
7188 25. When execve() failed while trying to run a command in a pipe transport,
7189 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7190 could be confused with a return value of 69 from the command itself. This
7191 has been changed to 127, the value the shell returns if it is asked to run
7192 a non-existent command. The wording for the related log line suggests a
7193 non-existent command as the problem.
7195 26. If received_header_text expands to an empty string, do not add a Received:
7196 header line to the message. (Well, it adds a token one on the spool, but
7197 marks it "old" so that it doesn't get used or transmitted.)
7199 27. Installed eximstats 1.28 (addition of -nt option).
7201 28. There was no check for failure on the call to getsockname() in the daemon
7202 code. This can fail if there is a shortage of resources on the system, with
7203 ENOMEM, for example. A temporary error is now given on failure.
7205 29. Contrary to the C standard, it seems that in some environments, the
7206 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7207 program. Exim now does this explicitly; it affects the formatting of
7208 timestamps using strftime().
7210 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7211 complaints. I've now initialized all the variables, to avoid this.
7213 32. Header lines added by a system filter were not being "seen" during
7214 transport-time rewrites.
7216 33. The info_callback() function passed to OpenSSL is set up with type void
7217 (*)(SSL *, int, int), as described somewhere. However, when calling the
7218 function (actually a macro) that sets it up, the type void(*)() is
7219 expected. I've put in a cast to prevent warnings from picky compilers.
7221 34. If a DNS black list lookup found a CNAME record, but there were no A
7222 records associated with the domain it pointed at, Exim crashed.
7224 35. If a DNS black list lookup returned more than one A record, Exim ignored
7225 all but the first. It now scans all returned addresses if a particular IP
7226 value is being sought. In this situation, the contents of the
7227 $dnslist_value variable are a list of all the addresses, separated by a
7230 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7231 to accept a host name and all its aliases if the forward lookup for any of
7232 them yielded the IP address of the incoming connection. Now it accepts only
7233 those names whose forward lookup yields the correct IP address. Any other
7234 names are discarded. This closes a loophole whereby a rogue DNS
7235 administrator could create reverse DNS records to break through a
7236 wildcarded host restriction in an ACL.
7238 37. If a user filter or a system filter that ran in a subprocess used any of
7239 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7240 the wrong values were passed to the pipe command ($thisaddress had the
7241 value of $0, $0 had the value of $1, etc). This bug was introduced by
7242 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7244 38. Improved the line breaking for long SMTP error messages from ACLs.
7245 Previously, if there was no break point between 40 and 75 characters, Exim
7246 left the rest of the message alone. Two changes have been made: (a) I've
7247 reduced the minimum length to 35 characters; (b) if it can't find a break
7248 point between 35 and 75 characters, it looks ahead and uses the first one
7249 that it finds. This may give the occasional overlong line, but at least the
7250 remaining text gets split now.
7252 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7253 file descriptors might be low, and that setting 1000 would always raise it.
7254 It turns out that in some environments, the limit is already over 1000 and
7255 that lowering it causes trouble. So now Exim takes care not to decrease it.
7257 40. When delivering a message, the value of $return_path is set to $sender_
7258 address at the start of routing (routers may change the value). By an
7259 oversight, this default was not being set up when an address was tested by
7260 -bt or -bv, which affected the outcome if any router or filter referred to
7263 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7264 log only when "message" or "log_message" are set. However, if one of the
7265 conditions was an address verification, or a call to a nested ACL, the
7266 messages generated by the underlying test were being passed through. This
7267 no longer happens. The underlying message is available in $acl_verify_
7268 message for both "message" and "log_message" expansions, so it can be
7269 passed through if needed.
7271 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7272 new expansion $bh_ to give the encoded byte string without charset
7273 translation. Translation happens only if iconv() is available; HAVE_ICONV
7274 indicates this at build time. HEADERS_CHARSET gives the charset to
7275 translate to; headers_charset can change it in the configuration, and
7276 "headers charset" can change it in an individual filter file.
7278 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7279 that creates RFC 2047 encoded "words" labels them as that charset instead
7280 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7281 expansion operator; (ii) when Exim creates a From: line for a local
7282 message; (iii) when a header line is rewritten to include a "phrase" part.
7284 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7285 buggy, causing it to skip the first lines of messages whose message ID
7286 ended in 'D'. This would not have bitten before Exim release 4.14, because
7287 message IDs were unlikely to end in 'D' before then. The effect was to have
7288 incorrect size information for certain domains.
7290 45. #include "config.h" was missing at the start of the crypt16.c module. This
7291 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7294 46. If there was a timeout during a "random" callout check, Exim treated it as
7295 a failure of the random address, and carried on sending RSET and the real
7296 address. If the delay was just some slowness somewhere, the response to the
7297 original RCPT would be taken as a response to RSET and so on, causing
7298 mayhem of various kinds.
7300 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7301 when I implemented it. It didn't allow for the fact that some option values
7302 may legitimately be negative (e.g. size_addition), and it didn't even do
7303 the right test for positive values.
7305 48. Domain names in DNS records are case-independent. Exim always looks them up
7306 in lower case. Some resolvers return domain names in exactly the case they
7307 appear in the zone file, that is, they may contain uppercase letters. Not
7308 all resolvers do this - some return always lower case. Exim was treating a
7309 change of case by a resolver as a change of domain, similar to a widening
7310 of a domain abbreviation. This triggered its re-routing code and so it was
7311 trying to route what was effectively the same domain again. This normally
7312 caused routing to fail (because the router wouldn't handle the domain
7313 twice). Now Exim checks for this case specially, and just changes the
7314 casing of the domain that it ultimately uses when it transmits the message
7317 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7320 50. If a filter generated a file delivery with a non-absolute name (possible if
7321 no home directory exists for the router), the forbid_file option was not
7324 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7325 the existing equality matching.
7327 52. Exim was using ints instead of ino_t variables in some places where it was
7328 dealing with inode numbers.
7330 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7331 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7332 called TMPDIR, and if it finds it is different, it changes its value.
7334 54. The smtp_printf() function is now made available to local_scan() so
7335 additional output lines can be written before returning. There is also an
7336 smtp_fflush() function to enable the detection of a dropped connection.
7337 The variables smtp_input and smtp_batched_input are exported to
7340 55. Changed the default runtime configuration: the message "Unknown user"
7341 has been removed from the ACL, and instead placed on the localuser router,
7342 using the cannot_route_message feature. This means that any verification
7343 failures that generate their own messages won't get overridden. Similarly,
7344 the "Unrouteable address" message that was in the ACL for unverifiable
7345 relay addresses has also been removed.
7347 56. Added hosts_avoid_esmtp to the smtp transport.
7349 57. The exicyclog script was not checking for the esoteric option
7350 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7351 will work only if exicyclog is run under the appropriate euid.
7353 58. Following a discussion on the list, the rules by which Exim recognises line
7354 endings on incoming messages have been changed. The -dropcr and drop_cr
7355 options are now no-ops, retained only for backwards compatibility. The
7356 following line terminators are recognized: LF CRLF CR. However, special
7357 processing applies to CR:
7359 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7360 nor a local message in the state where . is a terminator.
7362 (ii) If a bare CR is encountered in a header line, an extra space is added
7363 after the line terminator so as not to end the header. The reasoning
7364 behind this is that bare CRs in header lines are most likely either
7365 to be mistakes, or people trying to play silly games.
7367 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7368 was being incorrectly given as 18 bytes larger than it should have been.
7369 This is a VOB (very old bug).
7371 60. This may never have affected anything current, but just in case it has:
7372 When the local host is found other than at the start of a list of hosts,
7373 the local host, those with the same MX, and any that follow, are discarded.
7374 When the list in question was part of a longer list of hosts, the following
7375 hosts (not currently being processed) were also being discarded. This no
7376 longer happens. I'm not sure if this situation could ever has previously
7379 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7382 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7383 header. This is something that is recommended in a new Internet Draft, and
7384 is something that is documented as being done by Sendmail. There are two
7385 possible values. For messages generated by the autoreply transport, Exim
7388 Auto-Submitted: auto-replied
7390 whereas for all other generated messages (e.g. bounces) it adds
7392 Auto-Submitted: auto-generated
7394 63. The "personal" condition in filters now includes a test for the
7395 Auto-Submitted: header. If it contains the string "auto-" the message it
7396 not considered personal.
7398 64. Added rcpt_include_affixes as a generic transport option.
7400 65. Added queue_only_override (default true).
7402 66. Added the syslog_duplication option.
7404 67. If what should have been the first header line of a message consisted of
7405 a space followed by a colon, Exim was mis-interpreting it as a header line.
7406 It isn't of course - it is syntactically invalid and should therefore be
7407 treated as the start of the message body. The misbehaviour could have
7408 caused a number of strange effects, including loss of data in subsequent
7409 header lines, and spool format errors.
7411 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7412 client host had authenticated. This control can now be exercised by an ACL
7413 for more flexibility.
7415 69. By default, callouts do not happen when testing with -bh. There is now a
7416 variant, -bhc, which does actually run the callout code, including
7417 consulting and updating the callout cache.
7419 70. Added support for saslauthd authentication, courtesy of Alexander
7422 71. If statvfs() failed on the spool or log directories while checking their
7423 size for availability, Exim confusingly gave the error "space shortage".
7424 Furthermore, in debugging mode it crashed with a floating point exception.
7425 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7426 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7427 serious problem, Exim now writes to the main and panic logs when this
7428 happens, with details of the failure. It then refuses to accept the
7429 incoming message, giving the message "spool directory problem" or "log
7430 directory problem" with a 421 code for SMTP messages.
7432 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7433 0, 1, and 2 exist, because some OS complain for execs without them (see
7434 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7435 descriptors. However, the code omitted to check that the open succeeded,
7436 causing mysterious errors if for some reason the permissions on /dev/null
7437 got screwed. Now Exim writes a message to the main and panic logs, and
7438 bombs out if it can't open /dev/null.
7440 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7441 interact so that it is all more flexible. It is supposed to remain
7442 backwards compatible. Also added extra_local_interfaces.
7444 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7445 to bomb out with an assertion failure - to the client this appears as a
7446 connection drop. This problem occurs in the part of the code that was taken
7447 from the Samba project. Fortunately, the assertion is in a very simple
7448 function, so I have fixed this by reproducing the function inline in the
7449 one place where it is called, and arranging for authentication to fail
7450 instead of killing the process with assert().
7452 75. The SPA client code was not working when the server requested OEM rather
7453 than Unicode encoding.
7455 76. Added code to make require_files with a specific uid setting more usable in
7456 the case where statting the file as root fails - usually a non-root-mounted
7457 NFS file system. When this happens and the failure is EACCES, Exim now
7458 forks a subprocess and does the per-uid checking as the relevant uid.
7460 77. Added process_log_path.
7462 78. If log_file_path was not explicitly set, a setting of check_log_space or
7463 check_log_inodes was ignored.
7465 79. If a space check for the spool or log partitions fails, the incident is now
7466 logged. Of course, in the latter case the data may get lost...
7468 80. Added the %p formatting code to string_format() so that it can be used to
7469 print addresses in debug_print(). Adjusted all the address printing in the
7470 debugging in store.c to use %p rather than %d.
7472 81. There was a concern that a line of code in smtp_in.c could overflow a
7473 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7474 initially expressed, the concern was not well-founded, because trailing
7475 spaces are removed early. However, if the trailing spaces were followed by
7476 a NULL, they did not get removed, so the overflow was possible. Two fixes
7479 (a) I re-wrote the offending code in a cleaner fashion.
7480 (b) If an incoming SMTP command contains a NULL character, it is rejected
7483 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7484 runs initgroups(), so that if the Exim user is in any additional groups,
7485 they will be used during message reception.
7491 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob