1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platform supports and has the facility enabled, it will
125 be requested on all coneections.
127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
130 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
132 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
133 success-DSN messages. Previously the From: header was always the default
134 one for these; the option was ignored.
136 JH/28 Fix the timeout on smtp response to apply to the whole response.
137 Previously it was reset for every read, so a teergrubing peer sending
138 single bytes within the time limit could extend the connection for a
139 long time. Credit to Qualsys Security Advisory Team for the discovery.
141 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
142 delivery address, which leaked information of the results of local
143 forwarding. Change to the original envelope recipient address, per
146 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
147 requested. Previously not bounce was generated and a log entry of
148 error ignored was made.
150 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
152 JH/32 Introduce a general tainting mechanism for values read from the input
153 channel, and values derived from them. Refuse to expand any tainted
154 values, to catch one form of exploit.
160 JH/01 Remove code calling the customisable local_scan function, unless a new
161 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
163 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
164 non-signal-safe functions being used.
166 JH/03 Bug 2269: When presented with a received message having a stupidly large
167 number of DKIM-Signature headers, disable DKIM verification to avoid
168 a resource-consumption attack. The limit is set at twenty.
170 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
171 report of oldest_pass in ${authres } in consequence, and separate out
172 some descriptions of reasons for verification fail.
174 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
175 files in the spool were present and unlocked. A queue-runner could spot
176 them, resulting in a duplicate delivery. Fix that by doing the unlock
177 after the unlink. Investigation by Tim Stewart. Take the opportunity to
178 add more error-checking on spoolfile handling while that code is being
181 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
182 No known attacks, no CVE, this is defensive hardening.
184 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
185 a queue-runner could start a delivery while other operations were ongoing.
186 Cutthrough delivery was a common victim, resulting in duplicate delivery.
187 Found and investigated by Tim Stewart. Fix by using the open message data
188 file handle rather than opening another, and not locally closing it (which
189 releases a lock) for that case, while creating the temporary .eml format
190 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
192 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
193 $sender_verify_failure/$recipient_verify_failure to "random".
195 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
198 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
199 Previously this would segfault.
201 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
204 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
205 like zero, since the resolver should be doing this for us, But we need one
206 as a CNAME but no MX presence gets the CNAME returned; we need to check
207 that doesn't point to an MX to declare it "no MX returned" rather than
208 "error, loop". A new main option is added so the older capability of
209 following some limited number of chain links is maintained.
211 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
213 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
214 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
215 now-deprecated earlier definitions used only specified the range up to TLS
216 1.2 (in the older-version library docs).
218 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
220 JH/15 Rework TLS client-side context management. Stop using a global, and
221 explicitly pass a context around. This enables future use of TLS for
222 connections to service-daemons (eg. malware scanning) while a client smtp
223 connection is using TLS; with cutthrough connections this is quite likely.
225 JH/16 Fix ARC verification to do AS checks in reverse order.
227 JH/17 Support a "tls" option on the ${readsocket } expansion item.
229 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
230 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
231 Previously the "utf8" would be re-prepended for every additional message.
233 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
234 Previously thery were accepted, resulting in issues when attempting to
235 forward messages to a non-supporting MTA.
237 PP/02 Let -n work with printing macros too, not just options.
239 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
240 one parent address was copied, and bogus data was used at delivery-logging
241 time. Either a crash (after delivery) or bogus log data could result.
242 Discovery and analysis by Tim Stewart.
244 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
245 Previously if the string ended mid-character, we did not insert the
246 promised '?' replacement.
248 PP/04 Documentation: current string operators work on bytes, not codepoints.
250 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
251 should pack well giving a smaller memory footprint so better caching and
252 therefore performance. Group the declarations where this can't be done so
253 that the byte-sized flag variables are not interspersed among pointer
254 variables, giving a better chance of good packing by the compiler.
256 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
257 non-null, to avoid issues with sites running BATV. Previously reports were
258 sent with an empty envelope sender so looked like bounces.
260 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
261 The ignore_error flag wasn't being returned from the filter subprocess so
262 was not set for later routers. Investigation and fix by Matthias Kurz.
264 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
265 and a msg:complete for the whole, when a message is manually removed using
266 -Mrm. Developement by Matthias Kurz, hacked on by JH.
268 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
269 a "Gnu special" function, asprintf() in the DB utility binary builds; I
270 hope that is portable enough.
272 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
273 requiring a known-CA anchor certificate; make it now rely entirely on the
274 TLSA as an anchor. Checking the name on the leaf cert against the name
275 on the A-record for the host is still done for TA (but not for EE mode).
277 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
278 included in delivery lines for non-proxied connections, when compiled with
279 SUPPORT_SOCKS and running with proxy logging enabled.
281 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
282 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
283 move the existing event to fire before the normal logging of message
284 failure so that custom logging is bracketed by normal logging.
286 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
287 msg:fail:internal event. Developement by Matthias Kurz.
289 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
290 far too small for todays use of crypto signatures stored there. Go all
291 the way to the max DNS message size of 64kB, even though this might be
292 overmuch for IOT constrained device use.
294 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
295 copy a string over itself. The library routine is documented as not
296 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
298 JH/32 For main options check_spool_space and check_inode_space, where the
299 platform supports 64b integers, support more than the previous 2^31 kB
300 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
301 the previous G, M, k.
303 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
304 $authenticated_fail_id variable on authentication failure. Previously
307 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
308 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
309 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
310 GNUTLS_SEC_PARAM_MEDIUM.
312 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
313 side. Previously we would continue as if no SNI had been received.
315 JH/36 Harden the handling of string-lists. When a list consisted of a sole
316 "<" character, which should be a list-separator specification, we walked
317 off past the nul-terimation.
319 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
320 causes) even when the retry time is not yet met. Previously they were
321 not, meaning that when (say) an account was over-quota and temp-rejecting,
322 and multiple senders' messages were queued, only one sender would get
323 notified on each configured delay_warning cycle.
325 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
327 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
328 cipher-suites, an error can be left on the stack even for a succeeding
329 accept; this results in impossible error messages when a later operation
332 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
333 return error codes indicating retry. Under TLS1.3 this becomes required.
335 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
336 it only wrote the new authenticators, resulting in a lack of tracking of
337 peer changes of ESMTP extensions until the next cache flush.
339 JH/41 Fix the loop reading a message header line to check for integer overflow,
340 and more-often against header_maxsize. Previously a crafted message could
341 induce a crash of the recive process; now the message is cleanly rejected.
343 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
344 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
350 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
351 When redis_servers is set to a list of > 1 element, and the Redis servers
352 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
353 case of MOVED into a DEFER case instead, thus moving the query onto the
354 next server in the list. For a cluster of N elements, all N servers must
355 be defined in redis_servers.
357 GF/02 Catch and remove uninitialized value warning in exiqsumm
358 Check for existence of @ARGV before looking at $ARGV[0]
360 JH/01 Replace the store_release() internal interface with store_newblock(),
361 which internalises the check required to safely use the old one, plus
362 the allocate and data copy operations duplicated in both (!) of the
363 extant use locations.
365 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
366 modifier. This matches the restriction on the commandline.
368 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
369 Previously only the last row was returned.
371 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
372 we assumed that tags in the header were well-formed, and parsed the
373 element content after inspecting only the first char of the tag.
374 Assumptions at that stage could crash the receive process on malformed
377 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
378 While running the DKIM ACL we operate on the Permanent memory pool so that
379 variables created with "set" persist to the DATA ACL. Also (at any time)
380 DNS lookups that fail create cache records using the Permanent pool. But
381 expansions release any allocations made on the current pool - so a dnsdb
382 lookup expansion done in the DKIM ACL releases the memory used for the
383 DNS negative-cache, and bad things result. Solution is to switch to the
384 Main pool for expansions.
385 While we're in that code, add checks on the DNS cache during store_reset,
386 active in the testsuite.
387 Problem spotted, and debugging aided, by Wolfgang Breyha.
389 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
390 When none of the hosts presented to a transport match an already-open
391 connection, close it and proceed with the list. Previously we would
392 queue the message. Spotted by Lena with Yahoo, probably involving
395 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
396 Previously a spurious "250 OK id=" response was appended to the proper
399 JH/08 The "support for" informational output now, which built with Content
400 Scanning support, has a line for the malware scanner interfaces compiled
401 in. Interface can be individually included or not at build time.
403 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
404 by the template makefile "src/EDITME". The "STREAM" support for an older
405 ClamAV interface method is removed.
407 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
408 rows affected is given instead).
410 JH/11 The runtime Berkeley DB library version is now additionally output by
411 "exim -d -bV". Previously only the compile-time version was shown.
413 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
414 SMTP connection. Previously, when one had more receipients than the
415 first, an abortive onward connection was made. Move to full support for
416 multiple onward connections in sequence, handling cutthrough connection
417 for all multi-message initiating connections.
419 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
420 routers. Previously, a multi-recipient message would fail to match the
421 onward-connection opened for the first recipient, and cause its closure.
423 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
424 a timeout on read on a GnuTLS initiating connection, resulting in the
425 initiating connection being dropped. This mattered most when the callout
426 was marked defer_ok. Fix to keep the two timeout-detection methods
429 JH/15 Relax results from ACL control request to enable cutthrough, in
430 unsupported situations, from error to silently (except under debug)
431 ignoring. This covers use with PRDR, frozen messages, queue-only and
434 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
436 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
437 metadata, resulting in a crash in free().
439 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
440 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
441 Broken also in d185889f4, with init system revamp.
443 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
444 Previously we abruptly closed the connection after reading a malware-
445 found indication; now we go on to read the "scan ok" response line,
448 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
449 ACL. Previously, a crash would result.
451 JH/19 Speed up macro lookups during configuration file read, by skipping non-
452 macro text after a replacement (previously it was only once per line) and
453 by skipping builtin macros when searching for an uppercase lead character.
455 JH/20 DANE support moved from Experimental to mainline. The Makefile control
456 for the build is renamed.
458 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
459 was allocated for every new TLS startup, meaning one per message. Fix
460 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
462 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
463 reported the original. Fix to report (as far as possible) the ACL
464 result replacing the original.
466 JH/23 Fix memory leak during multi-message connections using STARTTLS under
467 OpenSSL. Certificate information is loaded for every new TLS startup,
468 and the resources needed to be freed.
470 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
472 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
475 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
476 DATA response info to the (existing) per-recipient response info for
477 the "C=" log element. It can have useful tracking info from the
478 destination system. Patch from Simon Arlott.
480 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
481 length value. Previously this would segfault.
483 HS/02 Support Avast multiline protoocol, this allows passing flags to
484 newer versions of the scanner.
486 JH/28 Ensure that variables possibly set during message acceptance are marked
487 dead before release of memory in the daemon loop. This stops complaints
488 about them when the debug_store option is enabled. Discovered specifically
489 for sender_rate_period, but applies to a whole set of variables.
490 Do the same for the queue-runner and queue-list loops, for variables set
491 from spool message files. Do the same for the SMTP per-message loop, for
492 certain variables indirectly set in ACL operations.
494 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
495 as a multi-recipient message from a mailinglist manager). The coding had
496 an arbitrary cutoff number of characters while checking for more input;
497 enforced by writing a NUL into the buffer. This corrupted long / fast
498 input. The problem was exposed more widely when more pipelineing of SMTP
499 responses was introduced, and one Exim system was feeding another.
500 The symptom is log complaints of SMTP syntax error (NUL chars) on the
501 receiving system, and refused recipients seen by the sending system
502 (propating to people being dropped from mailing lists).
503 Discovered and pinpointed by David Carter.
505 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
506 replaced by the ${authresults } expansion.
508 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
510 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
511 allows proper process termination in container environments.
513 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
514 Previously the "final dot" had a newline after it; ensure it is CR,LF.
516 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
517 and "err_perm", deprecated since 4.83 when the RFC-defined words
518 "temperror" and "permerror" were introduced.
520 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
521 transport-filters or DKIM-signing. The restriction was lost in the
522 consolidation of verify-callout and delivery SMTP handling.
523 Extend the restriction to also cover ARC-signing.
525 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
526 in defer=pass mode supply a 450 to the initiator. Previously the message
529 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
530 tls_require_ciphers is used as before.
532 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
533 "pass_unscanned". Only tmpfails from the scanner are written to
534 the paniclog, as they may require admin intervention (permission
535 denied, license issues). Other scanner errors (like decompression
536 bombs) do not cause a paniclog entry.
538 JH/36 Fix reinitialisation of DKIM logging variable between messages.
539 Previously it was possible to log spurious information in receive log
542 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
543 triggered odd behaviour from Outlook Express clients.
545 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
548 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
549 since the IETF WG has not yet settled on that versus the original
550 "bare" representation.
552 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
553 Previously the millisecond value corrupted the output.
554 Fix also for syslog_pid=no and log_selector +pid, for which the pid
555 corrupted the output.
561 JH/01 Rework error string handling in TLS interface so that the caller in
562 more cases is responsible for logging. This permits library-sourced
563 string to be attached to addresses during delivery, and collapses
564 pairs of long lines into single ones.
566 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
567 during configuration. Wildcards are allowed and expanded.
569 JH/02 Rework error string handling in DKIM to pass more info back to callers.
570 This permits better logging.
572 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
573 do not close it down and have the child transport start it up again on
574 the passed-on TCP connection. Instead, proxy the child (and any
575 subsequent ones) for TLS via a unix-domain socket channel. Logging is
576 affected: the continued delivery log lines do not have any DNSSEC, TLS
577 Certificate or OCSP information. TLS cipher information is still logged.
579 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
580 identical IP addresses on different listening ports. Will also affect
583 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
584 add noisy ifdef guards to special-case this sillyness.
585 Patch from Bernd Kuhls.
587 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
588 than 255 are no longer allowed.
590 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
591 Disable the session-cache too, which might reduce our load. Since we
592 currrectly use a new context for every connection, both as server and
593 client, there is no benefit for these.
594 GnuTLS appears to not support tickets server-side by default (we don't
595 call gnutls_session_ticket_enable_server()) but client side is enabled
596 by default on recent versions (3.1.3 +) unless the PFS priority string
599 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
600 <https://reproducible-builds.org/specs/source-date-epoch/>.
602 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
603 the check for any unsuccessful recipients did not notice the limit, and
604 erroneously found still-pending ones.
606 JH/08 Pipeline CHUNKING command and data together, on kernels that support
607 MSG_MORE. Only in-clear (not on TLS connections).
609 JH/09 Avoid using a temporary file during transport using dkim. Unless a
610 transport-filter is involved we can buffer the headers in memory for
611 creating the signature, and read the spool data file once for the
612 signature and again for transmission.
614 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
615 4.77 as the kernel support then wasn't solid, having issues in 64bit
616 mode. Now, it's been long enough. Add support for FreeBSD also.
618 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
619 case where the routing stage had gathered several addresses to send to
620 a host before calling the transport for the first, we previously failed
621 to close down TLS in the old transport process before passing the TCP
622 connection to the new process. The new one sent a STARTTLS command
623 which naturally failed, giving a failed delivery and bloating the retry
624 database. Investigation and fix prototype from Wolfgang Breyha.
626 JH/12 Fix check on SMTP command input synchronisation. Previously there were
627 false-negatives in the check that the sender had not preempted a response
628 or prompt from Exim (running as a server), due to that code's lack of
629 awareness of the SMTP input buffering.
631 PP/04 Add commandline_checks_require_admin option.
632 Exim drops privileges sanely, various checks such as -be aren't a
633 security problem, as long as you trust local users with access to their
634 own account. When invoked by services which pass untrusted data to
635 Exim, this might be an issue. Set this option in main configuration
636 AND make fixes to the calling application, such as using `--` to stop
639 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
640 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
641 responses to those, into a single TLS record each way (this usually means
642 a single packet). As a side issue, smtp_enforce_sync now works on TLS
645 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
646 affects you only if you're dancing at the edge of the param size limits.
647 If you are, and this message makes sense to you, then: raise the
648 configured limit or use OpenSSL 1.1. Nothing we can do for older
651 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
652 cmdline element to get the documented default one. Previously it was
655 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
656 are unsuable for non-message-specific reasons - eg. connection timeout,
657 banner-time rejection.
659 JH/16 Fix logging of delivery remote port, when specified by router, under
662 PP/06 Repair manualroute's ability to take options in any order, even if one
663 is the name of a transport.
666 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
668 JH/17 Change the list-building routines interface to use the expanding-string
669 triplet model, for better allocation and copying behaviour.
671 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
672 Previously it was constructed the first time a possibly-matching string
673 was met in the configuration file input during startup; now it is done
676 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
677 compatible one, to avoid the (poorly documented) possibility of a config
678 file in the working directory redirecting the DB files, possibly correpting
679 some existing file. CVE-2017-10140 assigned for BDB.
681 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
682 cache-hot. Previously, although the result was properly cached, the
683 initial verify call returned a defer.
685 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
686 the main verify for receipient in uncached-mode.
688 JH/22 Retire historical build files to an "unsupported" subdir. These are
689 defined as "ones for which we have no current evidence of testing".
691 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
692 if present. Previously it was ignored.
694 JH/24 Start using specified-initialisers in C structure init coding. This is
695 a C99 feature (it's 2017, so now considered safe).
697 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
698 if was a fixed-sized field and bitmask ops via macros; it is now more
701 PP/07 GitHub PR 56: Apply MariaDB build fix.
702 Patch provided by Jaroslav Škarvada.
704 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
705 during Coverity cleanups [4.87 JH/47]
706 Diagnosis and fix provided by Michael Fischer v. Mollard.
708 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
709 the right size to place the terminating semicolon on its own folded
710 line, the header hash was calculated to an incorrect value thanks to
711 the (relaxed) space the fold became.
713 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
714 and confused the parent.
716 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
717 which could crash as a result. This could lead to undeliverable messages.
719 JH/28 Logging: "next input sent too soon" now shows where input was truncated
722 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
723 matters on fast-turnover and PID-randomising systems, which were getting
724 out-of-order delivery.
726 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
727 a possibly-overlapping copy. The symptom was that "Remote host closed
728 connection in response to HELO" was logged instead of the actual 4xx
731 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
732 Previously only that bufferd was discarded, resulting in SYMTP command
735 JH/32 DKIM: when a message has multiple signatures matching an identity given
736 in dkim_verify_signers, run the dkim acl once for each. Previously only
737 one run was done. Bug 2189.
739 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
740 "panic the current process" to "deliberately defer". The panic log is
741 still written with the problem list name; the mail and reject logs now
742 get a temp-reject line for the message that was being handled, saying
743 something like "domains check lookup or other defer". The SMTP 451
744 message is still "Temporary local problem".
746 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
747 A crafted sequence of BDAT commands could result in in-use memory beeing
748 freed. CVE-2017-16943.
750 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
751 from SMTP input. Previously it was always done; now only done for DATA
752 and not BDAT commands. CVE-2017-16944.
754 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
755 to the message (such as an overlong header line). Previously this was
756 not done and we did not exit BDAT mode. Followon from the previous item
757 though a different problem.
763 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
764 than -2003 did; needs libidn2 in addition to libidn.
766 JH/02 The path option on a pipe transport is now expanded before use.
768 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
769 Patch provided by "Björn", documentation fix added too.
771 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
772 missing a wire-to-host endian conversion.
774 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
775 close after a BDAT command line could be taken as a following command,
776 giving a synch failure. Fix by only checking for synch immediately
777 before acknowledging the chunk.
779 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
780 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
781 macro. Patches provided by Josh Soref.
783 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
784 Previously we did not; the RFC seems ambiguous and VRFY is not listed
785 by IANA as a service extension. However, John Klensin suggests that we
788 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
789 the dkim code may be unix-mode line endings rather than smtp wire-format
790 CRLF, so prepend a CR to any bare LF.
792 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
793 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
795 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
796 the first were themselves being wrongly included in the feed into dkim
797 processing; with most chunk sizes in use this resulted in an incorrect
798 body hash calculated value.
800 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
801 DKIM signature block, for verification. Although advised against by
802 standards it is specifically not ruled illegal.
804 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
806 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
807 missing a body hash (the bh= tag).
809 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
810 It seems that HAProxy sends the Proxy Protocol information in clear and
811 only then does a TLS startup, so do the same.
813 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
814 TCP connections (such as for Spamd) unless the daemon successfully set
815 Fast Open mode on its listening sockets. This fixes breakage seen on
816 too-old kernels or those not configured for Fast Open, at the cost of
817 requiring both directions being enabled for TFO, and TFO never being used
818 by non-daemon-related Exim processes.
820 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
821 endings, at least on the first header line. Try to canonify any that get
822 past that check, despite the cost.
824 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
825 now limited to an arbitrary five deep, while parsing addresses with the
826 strip_excess_angle_brackets option enabled.
828 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
829 instead leave the unprompted TLS handshake in socket buffer for the
830 TLS library to consume.
832 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
834 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
836 JH/16 Drop variables when they go out of scope. Memory management drops a whole
837 region in one operation, for speed, and this leaves assigned pointers
838 dangling. Add checks run only under the testsuite which checks all
839 variables at a store-reset and panics on a dangling pointer; add code
840 explicitly nulling out all the variables discovered. Fixes one known
841 bug: a transport crash, where a dangling pointer for $sending_ip_address
842 originally assigned in a verify callout, is re-used.
844 PP/06 Drop '.' from @INC in various Perl scripts.
846 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
848 PP/08 Reduce a number of compilation warnings under clang; building with
849 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
850 should be warning-free.
852 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
854 HS/01 Fix portability problems introduced by PP/08 for platforms where
855 realloc(NULL) is not equivalent to malloc() [SunOS et al].
857 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
858 chunk. This allows us to accept broken chunked messages. We need a more
859 general solution here.
861 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
862 already-broken messages in the queue.
864 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
866 JH/19 Fix reference counting bug in routing-generated-address tracking.
872 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
873 supports it and a size is available (ie. the sending peer gave us one).
875 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
876 years of being deprecated). The replacements are the ACLs
877 acl_smtp_mime and acl_not_smtp_mime.
879 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
880 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
881 or even an in-clear connection were permitted. Now, if the host lookup
882 was dnssec and dane was requested then the host is only used if the
883 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
884 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
885 if one fails this test.
886 This means that a poorly-configured remote DNS will make it incommunicado;
887 but it protects against a DNS-interception attack on it.
889 JH/04 Bug 1810: make continued-use of an open smtp transport connection
890 non-noisy when a race steals the message being considered.
892 JH/05 If main configuration option tls_certificate is unset, generate a
893 self-signed certificate for inbound TLS connections.
895 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
896 in rewrites and routers.
898 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
899 and logged a warning sing 4.83; now they are a configuration file error.
901 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
902 (lacking @domain). Apply the same qualification processing as RCPT.
904 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
906 JH/10 Support ${sha256:} applied to a string (as well as the previous
909 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
910 a cutthrough deliver is pending, as we always want to make a connection.
911 This also avoids re-routing the message when later placing the cutthrough
912 connection after a verify cache hit.
913 Do not update it with the verify result either.
915 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
916 when routing results in more than one destination address.
918 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
919 signing (which inhibits the cutthrough capability). Previously only
920 the presence of an option was tested; now an expansion evaluating as
921 empty is permissible (obviously it should depend only on data available
922 when the cutthrough connection is made).
924 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
925 the relevant preceding SMTP command did not note the pipelining mode.
927 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
928 Previously they were not counted.
930 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
931 as one having no matching records. Previously we deferred the message
932 that needed the lookup.
934 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
935 distinguished as "(=".
937 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
938 for missing MX records. Previously it only worked for missing A records.
940 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
942 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
943 after the data-go-ahead and data-ack. Patch from Jason Betts.
945 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
946 even for a "none" policy. Patch from Tony Meyer.
948 JH/22 Fix continued use of a connection for further deliveries. If a port was
949 specified by a router, it must also match for the delivery to be
952 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
953 When one of the recipients of a message was unsuitable for the connection
954 (has no matching addresses), we lost track of needing to mark it
955 deferred. As a result mail would be lost.
957 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
959 JH/25 Decoding ACL controls is now done using a binary search; the source code
960 takes up less space and should be simpler to maintain. Merge the ACL
961 condition decode tables also, with similar effect.
963 JH/26 Fix problem with one_time used on a redirect router which returned the
964 parent address unchanged. A retry would see the parent address marked as
965 delivered, so not attempt the (identical) child. As a result mail would
968 JH/27 Fix a possible security hole, wherein a process operating with the Exim
969 UID can gain a root shell. Credit to http://www.halfdog.net/ for
970 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
973 JH/28 Enable {spool,log} filesystem space and inode checks as default.
974 Main config options check_{log,spool}_{inodes,space} are now
975 100 inodes, 10MB unless set otherwise in the configuration.
977 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
978 Previously it only applied to the main-section connection policy
981 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
983 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
984 by me. Added RFC7919 DH primes as an alternative.
986 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
987 are not in the system include path.
989 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
990 GnuTLS, when a session startup failed (eg because the client disconnected)
991 Exim did stdio operations after fclose. This was exposed by a recent
992 change which nulled out the file handle after the fclose.
994 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
995 signed directly by the cert-signing cert, rather than an intermediate
996 OCSP-signing cert. This is the model used by LetsEncrypt.
998 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1000 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1001 an incoming connection.
1003 HS/02 Bug 1802: Do not half-close the connection after sending a request
1006 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1007 fallback to "prime256v1".
1009 JH/34 SECURITY: Use proper copy of DATA command in error message.
1010 Could leak key material. Remotely exploitable. CVE-2016-9963.
1016 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1017 and 3.4.4 - once the server is enabled to respond to an OCSP request
1018 it does even when not requested, resulting in a stapling non-aware
1019 client dropping the TLS connection.
1021 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1022 support variable-length bit vectors. No functional change.
1024 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1025 The I= interface field on outgoing lines is now after the H= remote
1026 host field, same as incoming lines. There is a separate
1027 outgoing_interface log selector which allows you to disable the
1030 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1031 If not running log_selector +smtp_connection the mainlog would be held
1032 open indefinitely after a "too many connections" event, including to a
1033 deleted file after a log rotate. Leave the per net connection logging
1034 leaving it open for efficiency as that will be quickly detected by the
1035 check on the next write.
1037 HS/01 Bug 1671: Fix post transport crash.
1038 Processing the wait-<transport> messages could crash the delivery
1039 process if the message IDs didn't exist for some reason. When
1040 using 'split_spool_directory=yes' the construction of the spool
1041 file name failed already, exposing the same netto behaviour.
1043 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1044 mime_regex ACL conditions.
1046 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1047 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1048 to HELO, local diagnostic string.
1050 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1051 log line to debug. Even when configured with a tls authenticator many
1052 client connections are expected to not authenticate in this way, so
1053 an authenticate fail is not an error.
1055 HS/02 Add the Exim version string to the process info. This way exiwhat
1056 gives some more detail about the running daemon.
1058 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1059 matter for fast-change records such as DNSBLs.
1061 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1062 retry record, even if constant. There may be multiple transports with
1063 different interface settings and the retry behaviour needs to be kept
1066 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1068 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1070 JH/10 Bug 840: fix log_defer_output option of pipe transport
1072 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1073 hosts_randomize. This matters a lot when combined with mua_wrapper.
1075 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1076 ${quote_pgsql:<string>} operator.
1078 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1080 JH/14 Change header file name for freeradius-client. Relevant if compiling
1081 with Radius support; from the Gentoo tree and checked under Fedora.
1083 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1085 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1086 option transport_filter, meaning no filtering.
1088 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1090 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1091 defaults to "*" (all hosts). The variable is now available when not built
1092 with TLS, default unset, mainly to enable keeping the testsuite sane.
1093 If a server certificate is not supplied (via tls_certificate) an error is
1094 logged, and clients will find TLS connections fail on startup. Presumably
1095 they will retry in-clear.
1096 Packagers of Exim are strongly encouraged to create a server certificate
1097 at installation time.
1099 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1100 with the $config_file variable.
1102 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1103 in transport context, after the attempt, and per-recipient. The latter type
1104 is per host attempted. The event data is the error message, and the errno
1105 information encodes the lookup type (A vs. MX) used for the (first) host,
1106 and the trailing two digits of the smtp 4xx response.
1108 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1109 to write to mainlog (or rejectlog, paniclog) in the window between file
1110 creation and permissions/ownership being changed. Particularly affects
1111 installations where exicyclog is run as root, rather than exim user;
1112 result is that the running daemon panics and dies.
1114 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1116 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1117 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1118 "pri" and "weight". Note that the previous implicit priority given by the
1119 list order is no longer honoured.
1121 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1122 for DKIM processing.
1124 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1125 by defining SUPPORT_SOCKS.
1127 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1128 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1129 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1130 variables are renamed to proxy_{local,external}_{address,port}.
1132 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1133 for a build by defining SUPPORT_I18N
1135 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1136 of the query string, and make ${quote_redis:} do that quoting.
1138 JH/29 Move Events support from Experimental to mainline, enabled by default
1139 and removable for a build by defining DISABLE_EVENT.
1141 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1143 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1144 cached by the daemon.
1146 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1147 by defining LOOKUP_REDIS. The libhiredis library is required.
1149 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1150 keys are given for lookup.
1152 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1153 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1154 only supported when built with TLS support. The PolarSSL SHA routines
1155 are still used when the TLS library is too old for convenient support.
1157 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1158 openssl_options), for security. OpenSSL forces this from version 1.1.0
1159 server-side so match that on older versions.
1161 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1162 allocation for $value could be released as the expansion processing
1163 concluded, but leaving the global pointer active for it.
1165 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1166 and to use the domains and local_parts ACL conditions.
1168 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1169 incorrectly not doubled on cutthrough transmission, hence seen as a
1170 body-termination at the receiving system - resulting in truncated mails.
1171 Commonly the sender saw a TCP-level error, and retransmitted the message
1172 via the normal store-and-forward channel. This could result in duplicates
1173 received - but deduplicating mailstores were liable to retain only the
1174 initial truncated version.
1176 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1178 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1180 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1181 we're in there, support oversigning also; bug 1309.
1183 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1185 HS/04 Add support for keep_environment and add_environment options.
1187 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1188 either intentional arithmetic overflow during PRNG, or testing config-
1191 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1192 delivery resulted in actual delivery. Cancel cutthrough before DATA
1195 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1196 rejects a recipient: pass the reject to the originator.
1198 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1199 Many were false-positives and ignorable, but it's worth fixing the
1202 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1203 for the new environment-manipulation done at startup. Move the routines
1204 from being local to tls.c to being global via the os.c file.
1206 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1207 an extract embedded as result-arg for a map, the first arg for extract
1208 is unavailable so we cannot tell if this is a numbered or keyed
1209 extraction. Accept either.
1215 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1218 JH/02 The smtp transport option "multi_domain" is now expanded.
1220 JH/03 The smtp transport now requests PRDR by default, if the server offers
1223 JH/04 Certificate name checking on server certificates, when exim is a client,
1224 is now done by default. The transport option tls_verify_cert_hostnames
1225 can be used to disable this per-host. The build option
1226 EXPERIMENTAL_CERTNAMES is withdrawn.
1228 JH/05 The value of the tls_verify_certificates smtp transport and main options
1229 default to the word "system" to access the system default CA bundle.
1230 For GnuTLS, only version 3.0.20 or later.
1232 JH/06 Verification of the server certificate for a TLS connection is now tried
1233 (but not required) by default. The verification status is now logged by
1234 default, for both outbound TLS and client-certificate supplying inbound
1237 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1240 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1241 Status Notification (bounce) messages are now MIME format per RFC 3464.
1242 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1243 under the control of the dsn_advertise_hosts option, and routers may
1244 have a dsn_lasthop option.
1246 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1247 default, modifiable by a malware= option. The list separator for
1248 the options can now be changed in the usual way. Bug 68.
1250 JH/10 The smtp_receive_timeout main option is now expanded before use.
1252 JH/11 The incoming_interface log option now also enables logging of the
1253 local interface on delivery outgoing connections.
1255 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1256 if the interface and destination host and port all match.
1258 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1261 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1262 Patch from Andrew Lewis.
1264 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1265 now supports optional time-restrictions, weighting, and priority
1266 modifiers per server. Patch originally by <rommer@active.by>.
1268 JH/16 The spamd_address main option now supports a mixed list of local
1269 and remote servers. Remote servers can be IPv6 addresses, and
1270 specify a port-range.
1272 JH/17 Bug 68: The spamd_address main option now supports an optional
1273 timeout value per server.
1275 JH/18 Bug 1581: Router and transport options headers_add/remove can
1276 now have the list separator specified.
1278 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1281 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1284 JH/21 Support for the A6 type of dns record is withdrawn.
1286 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1287 rather than the verbs used.
1289 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1290 from 255 to 1024 chars.
1292 JH/24 Verification callouts now attempt to use TLS by default.
1294 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1295 are generic router options now. The defaults didn't change.
1297 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1298 Original patch from Alexander Shikoff, worked over by JH.
1300 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1301 files if ZCAT_COMMAND is not executable.
1303 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1305 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1307 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1308 Normally benign, it bites when the pair was led to by a CNAME;
1309 modern usage is to not canonicalize the domain to a CNAME target
1310 (and we were inconsistent anyway for A-only vs AAAA+A).
1312 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1314 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1315 when evaluating $sender_host_dnssec.
1317 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1318 $sender_helo_dnssec variable.
1320 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1322 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1324 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1326 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1327 documented as working, but never had. Support all but $spam_report.
1329 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1330 added for tls authenticator.
1332 HS/03 Add perl_taintmode main config option
1338 TL/01 When running the test suite, the README says that variables such as
1339 no_msglog_check are global and can be placed anywhere in a specific
1340 test's script, however it was observed that placement needed to be near
1341 the beginning for it to behave that way. Changed the runtest perl
1342 script to read through the entire script once to detect and set these
1343 variables, reset to the beginning of the script, and then run through
1344 the script parsing/test process like normal.
1346 TL/02 The BSD's have an arc4random API. One of the functions to induce
1347 adding randomness was arc4random_stir(), but it has been removed in
1348 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1349 function when detected.
1351 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1352 cause callback expansion.
1354 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1355 syntax errors in an expansion can be treated as a string instead of
1356 logging or causing an error, due to the internal use of bool_lax
1357 instead of bool when processing it.
1359 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1360 server certificates when making smtp deliveries.
1362 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1364 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1366 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1368 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1369 Merged patch from Sebastian Wiedenroth.
1371 JH/05 Fix results-pipe from transport process. Several recipients, combined
1372 with certificate use, exposed issues where response data items split
1373 over buffer boundaries were not parsed properly. This eventually
1374 resulted in duplicates being sent. This issue only became common enough
1375 to notice due to the introduction of connection certificate information,
1376 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1378 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1379 size buffer was used, resulting in syntax errors when an expansion
1382 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1383 version 3.3.6 or later.
1385 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1386 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1387 both become "event_action", the variables become $event_name, $event_data
1388 and $event_defer_errno. There is a new variable $verify_mode, usable in
1389 routers, transports and related events. The tls:cert event is now also
1390 raised for inbound connections, if the main configuration event_action
1393 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1394 early OCSP support, but no stapling (appears to be less than 1.0.0).
1396 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1397 server certificate names available under the smtp transport option
1398 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1401 JH/10 Time-related extraction expansions from certificates now use the main
1402 option "timezone" setting for output formatting, and are consistent
1403 between OpenSSL and GnuTLS compilations. Bug 1541.
1405 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1406 encoded parameter in the incoming message. Bug 1558.
1408 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1409 include certificate info, eximon was claiming there were spoolfile
1412 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1414 JH/14 Log delivery-related information more consistently, using the sequence
1415 "H=<name> [<ip>]" wherever possible.
1417 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1418 are problematic for Debian distribution, omit them from the release
1421 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1423 JH/16 Fix string representation of time values on 64bit time_t architectures.
1426 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1427 output list separator was used.
1432 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1433 checkers that were complaining about end of non-void function with no
1436 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1437 This was a regression introduced in 4.83 by another bugfix.
1439 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1441 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1442 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1448 TF/01 Correctly close the server side of TLS when forking for delivery.
1450 When a message was received over SMTP with TLS, Exim failed to clear up
1451 the incoming connection properly after forking off the child process to
1452 deliver the message. In some situations the subsequent outgoing
1453 delivery connection happened to have the same fd number as the incoming
1454 connection previously had. Exim would try to use TLS and fail, logging
1455 a "Bad file descriptor" error.
1457 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1458 utilities have not been installed.
1460 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1461 temporary space as the ACL may create new global variables.
1463 TL/01 LDAP support uses per connection or global context settings, depending
1464 upon the detected version of the libraries at build time.
1466 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1467 to extract and use the src ip:port in logging and expansions as if it
1468 were a direct connection from the outside internet. PPv2 support was
1469 updated based on HAProxy spec change in May 2014.
1471 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1473 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1474 Properly escape header and check for NULL return.
1476 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1479 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1481 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1482 characters in header names, implemented as a verify condition.
1483 Contributed by Michael Fischer v. Mollard.
1485 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1486 results permerror and temperror. Previous values are deprecated but
1487 still accepted. In a future release, err_perm and err_temp will be
1488 completely removed, which will be a backward incompatibility if the
1489 ACL tests for either of these two old results. Patch contributed by
1490 user bes-internal on the mailing list.
1492 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1494 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1495 selectors, in both main and reject logs.
1497 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1500 JH/07 Add malware type "sock" for talking to simple daemon.
1502 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1504 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1505 routers/transports under cutthrough routing.
1507 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1508 numbers. Touch up "bool" conditional to keep the same definition.
1510 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1512 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1514 JH/12 Expand items in router/transport headers_add or headers_remove lists
1515 individually rather than the list as a whole. Bug 1452.
1517 Required for reasonable handling of multiple headers_ options when
1518 they may be empty; requires that headers_remove items with embedded
1519 colons must have them doubled (or the list-separator changed).
1521 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1522 view the policy declared in the DMARC record. Currently, $dmarc_status
1523 is a combined value of both the record presence and the result of the
1526 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1528 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1529 dnslookup router and the smtp transport (applying to the forward
1532 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1533 of ldap servers used for a specific lookup. Patch provided by Heiko
1536 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1537 New variable $lookup_dnssec_authenticated for observability.
1539 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1540 Patch submitted by Lars Timman.
1542 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1544 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1545 Requires trusted mode and valid format message id, aborts otherwise.
1546 Patch contributed by Heiko Schlichting.
1548 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1549 certextract with support for various fields. Bug 1358.
1551 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1552 is requested by default, modifiable by smtp transport option
1555 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1556 operate on certificate variables to give certificate fingerprints
1557 Also new ${sha256:cert_variable}.
1559 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1561 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1562 Christian Aistleitner.
1564 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1566 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1567 file. Patch from Wolfgang Breyha.
1569 JH/25 Expand the coverage of the delivery $host and $host_address to
1570 client authenticators run in verify callout. Bug 1476.
1572 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1573 align with daemon_smtp_ports. Bug 72.
1575 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1576 support and error reporting did not work properly.
1578 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1579 and is readable. Patch from Andrew Colin Kissa.
1581 TL/14 Enhance documentation of ${run expansion and how it parses the
1582 commandline after expansion, particularly in the case when an
1583 unquoted variable expansion results in an empty value.
1585 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1587 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1588 by renaming away from it. Addresses GH issue 15, reported by
1591 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1593 TL/15 SECURITY: prevent double expansion in math comparison functions
1594 (can expand unsanitized data). Not remotely exploitable.
1601 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1603 PP/02 Make -n do something, by making it not do something.
1604 When combined with -bP, the name of an option is not output.
1606 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1609 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1610 $sender_host_name and config options to manage this, and basic check
1613 PP/05 DSCP support for outbound connections and control modifier for inbound.
1615 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1616 (Only plugin which currently uses this is kerberos4, which nobody should
1617 be using, but we should make it available and other future plugins might
1618 conceivably use it, even though it would break NAT; stuff *should* be
1619 using channel bindings instead).
1621 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1622 name; added for Sendmail compatibility; requires admin caller.
1623 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1624 just ignore it); requires trusted caller.
1625 Also parse but ignore: -Ac -Am -X<logfile>
1628 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1630 TL/02 Add +smtp_confirmation as a default logging option.
1632 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1633 Patch by Magnus Holmgren from 2007-02-20.
1635 TL/04 Bugzilla 1281 - Spec typo.
1636 Bugzilla 1283 - Spec typo.
1637 Bugzilla 1290 - Spec grammar fixes.
1639 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1641 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1643 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1644 mailing list by by Dmitry Isaikin.
1646 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1648 JH/02 Support "G" suffix to numbers in ${if comparisons.
1650 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1652 NM/01 Bugzilla 1197 - Spec typo
1653 Bugzilla 1196 - Spec examples corrections
1655 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1657 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1658 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1661 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1662 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1664 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1665 "acl {{name}{arg}...}", and optional args on acl condition
1668 JH/05 Permit multiple router/transport headers_add/remove lines.
1670 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1672 JH/07 Avoid using a waiting database for a single-message-only transport.
1673 Performance patch from Paul Fisher. Bugzilla 1262.
1675 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1678 JH/09 Add $headers_added variable, with content from use of ACL modifier
1679 add_header (but not yet added to the message). Bugzilla 199.
1681 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1682 Pulled from Bugzilla 817 by Wolfgang Breyha.
1684 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1686 (nb: this is the same fix as in Exim 4.80.1)
1688 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1691 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1694 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1696 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1697 advertise SMTP AUTH mechanism to us, instead of a generic
1698 protocol violation error. Also, make Exim more robust to bad
1699 data from the Dovecot auth socket.
1701 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1703 When a queue runner is handling a message, Exim first routes the
1704 recipient addresses, during which it prunes them based on the retry
1705 hints database. After that it attempts to deliver the message to
1706 any remaining recipients. It then updates the hints database using
1709 So if a recipient address works intermittently, it can get repeatedly
1710 deferred at routing time. The retry hints record remains fresh so the
1711 address never reaches the final cutoff time.
1713 This is a fairly common occurrence when a user is bumping up against
1714 their storage quota. Exim had some logic in its local delivery code
1715 to deal with this. However it did not apply to per-recipient defers
1716 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1718 This change adds a proper retry rule check during routing so that the
1719 final cutoff time is checked against the message's age. We only do
1720 this check if there is an address retry record and there is not a
1721 domain retry record; this implies that previous attempts to handle
1722 the address had the retry_use_local_parts option turned on. We use
1723 this as an approximation for the destination being like a local
1724 delivery, as in LMTP.
1726 I suspect this new check makes the old local delivery cutoff check
1727 redundant, but I have not verified this so I left the code in place.
1729 TF/02 Correct gecos expansion when From: is a prefix of the username.
1731 Test 0254 submits a message to Exim with the header
1735 When I ran the test suite under the user fanf2, Exim expanded
1736 the header to contain my full name, whereas it should have added
1737 a Resent-Sender: header. It erroneously treats any prefix of the
1738 username as equal to the username.
1740 This change corrects that bug.
1742 GF/01 DCC debug and logging tidyup
1743 Error conditions log to paniclog rather than rejectlog.
1744 Debug lines prefixed by "DCC: " to remove any ambiguity.
1746 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1748 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1749 Bug spotted by Jeremy Harris; was flawed since initial commit.
1750 Would have resulted in OCSP responses post-SNI triggering an Exim
1751 NULL dereference and crash.
1753 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1755 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1756 Bug detection, analysis and fix by Samuel Thibault.
1757 Bugzilla 1331, Debian bug #698092.
1759 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1761 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1762 Server implementation by Todd Lyons, client by JH.
1763 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1764 config variable "prdr_enable" controls whether the server
1765 advertises the facility. If the client requests PRDR a new
1766 acl_data_smtp_prdr ACL is called once for each recipient, after
1767 the body content is received and before the acl_smtp_data ACL.
1768 The client is controlled by both of: a hosts_try_prdr option
1769 on the smtp transport, and the server advertisement.
1770 Default client logging of deliveries and rejections involving
1771 PRDR are flagged with the string "PRDR".
1773 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1774 fclose(). Diagnosis by Todd Lyons.
1776 PP/17 Update configure.default to handle IPv6 localhost better.
1777 Patch by Alain Williams (plus minor tweaks).
1780 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1781 This is now consistent with GnuTLS, and is now documented: the
1782 previous undocumented portable approach to treating the option as
1783 unset was to force an expansion failure. That still works, and
1784 an empty string is now equivalent.
1786 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1787 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1788 not performing validation itself.
1790 PP/20 Added force_command boolean option to pipe transport.
1791 Patch from Nick Koston, of cPanel Inc.
1793 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1796 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1798 PP/21 Fix eximon continuous updating with timestamped log-files.
1799 Broken in a format-string cleanup in 4.80, missed when I repaired the
1800 other false fix of the same issue.
1801 Report and fix from Heiko Schlichting.
1804 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1805 Report from Prashanth Katuri.
1807 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1808 It's SecureTransport, so affects any MacOS clients which use the
1809 system-integrated TLS libraries, including email clients.
1811 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1812 using a MIME ACL for non-SMTP local injection.
1813 Report and assistance in diagnosis by Warren Baker.
1815 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1817 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1819 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1820 last id that failed so it may be referenced in subsequent ACL's.
1822 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1825 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1826 ldap library initialization, allowing self-signed CA's to be
1827 used. Also properly sets require_cert option later in code by
1828 using NULL (global ldap config) instead of ldap handle (per
1829 session). Bug diagnosis and testing by alxgomz.
1831 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1832 the src/util/ subdirectory.
1834 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1835 renamed to Transport Post Delivery Action by Jeremy Harris, as
1838 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1839 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1840 redis_servers = needs to be configured which will be used by the redis
1841 lookup. Patch from Warren Baker, of The Packet Hub.
1843 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1845 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1846 hostname or reverse DNS when processing a host list. Used suggestions
1847 from multiple comments on this bug.
1849 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1851 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1852 Missed a few lines, added it to make the runtest require no keyboard
1855 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1856 contains upper case chars. Make router use caseful_local_part.
1858 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1859 support when GnuTLS has been built with p11-kit.
1865 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1867 This, or similar/improved, will also be change PP/11 of 4.82.
1873 PP/01 Handle short writes when writing local log-files.
1874 In practice, only affects FreeBSD (8 onwards).
1875 Bugzilla 1053, with thanks to Dmitry Isaikin.
1877 NM/01 Bugzilla 949 - Documentation tweak
1879 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1882 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1884 PP/02 Implemented gsasl authenticator.
1886 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1888 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1889 `pkg-config foo` for cflags/libs.
1891 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1892 with rest of GSASL and with heimdal_gssapi.
1894 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1895 `pkg-config foo` for cflags/libs for the TLS implementation.
1897 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1898 properties get this fed in as external SSF. A number of robustness
1899 and debugging improvements to the cyrus_sasl authenticator.
1901 PP/08 cyrus_sasl server now expands the server_realm option.
1903 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1904 Patch by Jeremy Harris.
1906 PP/10 Added dbmjz lookup type.
1908 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1910 PP/12 MAIL args handles TAB as well as SP, for better interop with
1911 non-compliant senders.
1912 Analysis and variant patch by Todd Lyons.
1914 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1915 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1916 Patch from Dirk Mueller <dmueller@suse.com>
1918 PP/13 tls_peerdn now print-escaped for spool files.
1919 Observed some $tls_peerdn in wild which contained \n, which resulted
1920 in spool file corruption.
1922 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1923 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1924 or write after TLS renegotiation, which otherwise led to messages
1927 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1928 as a tracking header (ie: a signed header comes before the signature).
1929 Patch from Wolfgang Breyha.
1931 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1932 comma-sep list; embedded commas doubled.
1934 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1936 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1938 Report and patch from Dmitry Banschikov.
1940 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1941 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1942 failures. We appear to now support TLS1.1+ with Exim.
1944 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1945 lets Exim select keys and certificates based upon TLS SNI from client.
1946 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1947 before an outbound SMTP session. New log_selector, +tls_sni.
1949 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1950 NULL dereference. Report and patch from Alun Jones.
1952 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1953 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1954 Not seeing resolver debug output on NetBSD, but suspect this is a
1955 resolver implementation change.
1957 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1958 Left warnings. Added "eximon gdb" invocation mode.
1960 PP/21 Defaulting "accept_8bitmime" to true, not false.
1962 PP/22 Added -bw for inetd wait mode support.
1964 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1965 locate the relevant includes and libraries. Made this the default.
1967 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1968 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1970 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1971 This may cause build issues on older platforms.
1973 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1974 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1975 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1976 Added SNI support via GnuTLS too.
1977 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1979 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1981 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1982 Applied second patch from Janne, implementing suggestion to default
1983 multiple-strings-in-record handling to match SPF spec.
1985 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1987 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1988 read-only, out of scope).
1989 Patch from Wolfgang Breyha, report from Stuart Northfield.
1991 PP/29 Fix three issues highlighted by clang analyser static analysis.
1992 Only crash-plausible issue would require the Cambridge-specific
1993 iplookup router and a misconfiguration.
1994 Report from Marcin Mirosław.
1996 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1998 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1999 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2000 As part of this, removing so much warning spew let me fix some minor
2001 real issues in debug logging.
2003 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2004 assignment on my part. Fixed.
2006 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2007 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2008 Janne Snabb (who went above and beyond: thank you).
2010 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2011 string otherwise requires a connection and a bunch more work and it's
2012 relatively easy to get wrong. Should also expose TLS library linkage
2015 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2016 64-bit ${eval} (JH/03).
2018 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2019 GNU libc to support some of the 64-bit stuff, should not lead to
2020 conflicts. Defined before os.h is pulled in, so if a given platform
2021 needs to override this, it can.
2023 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2024 protection layer was required, which is not implemented.
2025 Bugzilla 1254, patch from Wolfgang Breyha.
2027 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2028 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2029 tls_dhparam take prime identifiers. Also unbreak combination of
2030 OpenSSL+DH_params+TLSSNI.
2032 PP/39 Disable SSLv2 by default in OpenSSL support.
2038 PP/01 Solaris build fix for Oracle's LDAP libraries.
2039 Bugzilla 1109, patch from Stephen Usher.
2041 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2043 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2046 TF/02 Fix a couple more cases where we did not log the error message
2047 when unlink() failed. See also change 4.74-TF/03.
2049 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2050 lock up or crash if it happened to be inside a call to libc when it
2051 got a SIGUSR1 from exiwhat.
2053 The SIGUSR1 handler appends the current process status to the process
2054 log which is later printed by exiwhat. It used to use the general
2055 purpose logging code to do this, but several functions it calls are
2056 not safe for signals.
2058 The new output code in the SIGUSR1 handler is specific to the process
2059 log, and simple enough that it's easy to inspect for signal safety.
2060 Removing some special cases also simplifies the general logging code.
2061 Removing the spurious timestamps from the process log simplifies
2064 TF/04 Improved ratelimit ACL condition.
2066 The /noupdate option has been deprecated in favour of /readonly which
2067 has clearer semantics. The /leaky, /strict, and /readonly update modes
2068 are mutually exclusive. The update mode is no longer included in the
2069 database key; it just determines when the database is updated. (This
2070 means that when you upgrade Exim will forget old rate measurements.)
2072 Exim now checks that the per_* options are used with an update mode that
2073 makes sense for the current ACL. For example, when Exim is processing a
2074 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2075 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2076 must specify per_mail/readonly. If you omit the update mode it defaults to
2077 /leaky where that makes sense (as before) or /readonly where required.
2079 The /noupdate option is now undocumented but still supported for
2080 backwards compatibility. It is equivalent to /readonly except that in
2081 ACLs where /readonly is required you may specify /leaky/noupdate or
2082 /strict/noupdate which are treated the same as /readonly.
2084 A useful new feature is the /count= option. This is a generalization
2085 of the per_byte option, so that you can measure the throughput of other
2086 aggregate values. For example, the per_byte option is now equivalent
2087 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2089 The per_rcpt option has been generalized using the /count= mechanism
2090 (though it's more complicated than the per_byte equivalence). When it is
2091 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2092 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2093 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2094 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2095 Note that using per_rcpt with a non-readonly update mode in more than
2096 one ACL will cause the recipients to be double-counted. (The per_mail
2097 and per_byte options don't have this problem.)
2099 The handling of very low rates has changed slightly. If the computed rate
2100 is less than the event's count (usually one) then this event is the first
2101 after a long gap. In this case the rate is set to the same as this event's
2102 count, so that the first message of a spam run is counted properly.
2104 The major new feature is a mechanism for counting the rate of unique
2105 events. The new per_addr option counts the number of different
2106 recipients that someone has sent messages to in the last time period. It
2107 behaves like per_rcpt if all the recipient addresses are different, but
2108 duplicate recipient addresses do not increase the measured rate. Like
2109 the /count= option this is a general mechanism, so the per_addr option
2110 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2111 example, measure the rate that a client uses different sender addresses
2112 with the options per_mail/unique=$sender_address. There are further
2113 details in the main documentation.
2115 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2117 TF/06 Removed a few PCRE remnants.
2119 TF/07 Automatically extract Exim's version number from tags in the git
2120 repository when doing development or release builds.
2122 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2123 Bugzilla 879. Patch from Paul Fisher.
2125 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2126 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2129 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2131 PP/05 Variable $av_failed, true if the AV scanner deferred.
2132 Bugzilla 1078. Patch from John Horne.
2134 PP/06 Stop make process more reliably on build failure.
2135 Bugzilla 1087. Patch from Heiko Schlittermann.
2137 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2138 Bugzilla 1089. Patch from Heiko Schlittermann.
2140 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2141 Bugzilla 1131. Patch from Holger Weiß.
2143 PP/09 Handle IPv6 addresses with SPF.
2144 Bugzilla 860. Patch from Wolfgang Breyha.
2146 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2148 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2151 PP/11 match_* no longer expand right-hand-side by default.
2152 New compile-time build option, EXPAND_LISTMATCH_RHS.
2153 New expansion conditions, "inlist", "inlisti".
2155 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2157 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2159 PP/14 fix log_write() format string regression from TF/03.
2160 Bugzilla 1152. Patch from Dmitry Isaikin.
2166 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2168 PP/02 Harmonised TLS library version reporting; only show if debugging.
2169 Layout now matches that introduced for other libraries in 4.74 PP/03.
2171 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2173 PP/04 New "dns_use_edns0" global option.
2175 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2178 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2179 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2181 TK/01 Updated PolarSSL code to 0.14.2.
2182 Bugzilla 1097. Patch from Andreas Metzler.
2184 PP/07 Catch divide-by-zero in ${eval:...}.
2185 Fixes bugzilla 1102.
2187 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2190 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2191 format-string attack -- SECURITY: remote arbitrary code execution.
2193 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2194 time unintentionally subject to list matching rules, letting the header
2195 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2196 arbitrary string expansion). This allowed for information disclosure.
2198 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2199 INT_MIN/-1 -- value coerced to INT_MAX.
2205 NM/01 Workaround for PCRE version dependency in version reporting
2208 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2209 This fixes portability to compilers other than gcc, notably
2210 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2212 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2213 makefiles for portability to HP-UX and POSIX correctness.
2215 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2216 Also via indented variable definition in the Makefile.
2217 (Debugging by Oliver Heesakkers).
2219 PP/02 Restore caching of spamd results with expanded spamd_address.
2220 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2222 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2223 Improves build reliability. Fix from: Frank Elsner
2225 NM/02 Fix wide character breakage in the rfc2047 coding
2226 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2228 NM/03 Allow underscore in dnslist lookups
2229 Fixes bug 1026. Patch from Graeme Fowler
2231 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2232 Code patches from Adam Ciarcinski of NetBSD.
2234 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2237 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2238 is logged, to avoid truncation. Patch from John Horne.
2240 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2241 Patch from Jakob Hirsch.
2243 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2244 SQL string expansion failure details.
2245 Patch from Andrey Oktyabrski.
2247 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2248 Patch from Simon Arlott.
2250 PP/09 New lookups functionality failed to compile on old gcc which rejects
2251 extern declarations in function scope.
2252 Patch from Oliver Fleischmann
2254 PP/10 Use sig_atomic_t for flags set from signal handlers.
2255 Check getgroups() return and improve debugging.
2256 Fixed developed for diagnosis in bug 927 (which turned out to be
2259 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2260 Patch from Mark Zealey.
2262 PP/12 Bugzilla 1056: Improved spamd server selection.
2263 Patch from Mark Zealey.
2265 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2266 Based on patch from Heiko Schlittermann.
2268 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2269 Patch from Uwe Doering, sign-off by Michael Haardt.
2271 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2272 variable declaration deep within a block. Bug and patch from
2275 PP/15 lookups-Makefile IRIX compatibility coercion.
2277 PP/16 Make DISABLE_DKIM build knob functional.
2279 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2280 Patch by Simon Arlott
2282 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2283 variable argument macros. Our copy now differs from upstream.
2289 TF/01 Failure to get a lock on a hints database can have serious
2290 consequences so log it to the panic log.
2292 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2293 controlled using the smtp_confirmation log selector.
2295 TF/03 Include the error message when we fail to unlink a spool file.
2297 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2298 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2299 for maintaining out-of-tree patches for some time.
2301 PP/01 Bugzilla 139: Documentation and portability issues.
2302 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2303 Handle per-OS dynamic-module compilation flags.
2305 PP/02 Let /dev/null have normal permissions.
2306 The 4.73 fixes were a little too stringent and complained about the
2307 permissions on /dev/null. Exempt it from some checks.
2308 Reported by Andreas M. Kirchwitz.
2310 PP/03 Report version information for many libraries, including
2311 Exim version information for dynamically loaded libraries. Created
2312 version.h, now support a version extension string for distributors
2313 who patch heavily. Dynamic module ABI change.
2315 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2316 privilege escalation vulnerability whereby the Exim run-time user
2317 can cause root to append content of the attacker's choosing to
2320 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2323 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2324 If dropping privileges for untrusted macros, we disabled normal logging
2325 on the basis that it would fail; for the Exim run-time user, this is not
2326 the case, and it resulted in successful deliveries going unlogged.
2327 Fixed. Reported by Andreas Metzler.
2333 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2334 only prepend for the Resent-* case. Fixes regression introduced in
2335 Exim 4.70 by NM/22 for Bugzilla 607.
2337 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2338 increasing numbers of administrators be bitten by this.
2340 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2342 PP/03 Bugzilla 994: added openssl_options main configuration option.
2344 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2346 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2348 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2350 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2351 without a peer certificate, leading to a segfault because of an
2352 assumption that peers always have certificates. Be a little more
2353 paranoid. Problem reported by Martin Tscholak.
2355 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2356 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2357 NB: ClamAV planning to remove STREAM in "middle of 2010".
2358 CL also introduces -bmalware, various -d+acl logging additions and
2359 more caution in buffer sizes.
2361 PP/09 Implemented reverse_ip expansion operator.
2363 PP/10 Bugzilla 937: provide a "debug" ACL control.
2365 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2367 PP/12 Bugzilla 973: Implement --version.
2369 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2371 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2373 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2375 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2376 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2377 ignore trailing whitespace.
2379 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2381 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2384 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2385 Notification from Dr Andrew Aitchison.
2387 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2388 ExtendedDetectionInfo response format.
2389 Notification from John Horne.
2391 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2394 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2395 XSL and documented dependency on system catalogs, with examples of how
2398 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2401 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2402 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2403 configuration file which is writeable by the Exim user or group.
2405 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2406 of configuration files to cover files specified with the -C option if
2407 they are going to be used with root privileges, not just the default
2410 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2411 option (effectively making it always true).
2413 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2414 files to be used while preserving root privileges.
2416 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2417 that rogue child processes cannot use them.
2419 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2420 run-time user, instead of root.
2422 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2423 Exim run-time user without dropping privileges.
2425 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2426 result string, instead of calling string_vformat() twice with the same
2429 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2430 for other users. Others should always drop root privileges if they use
2431 -C on the command line, even for a whitelisted configure file.
2433 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2435 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2441 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2442 $data_path, and $header_path variables; fixed documentation bugs and
2445 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2446 exipick to access non-standard spools, including the "frozen" queue
2449 NM/01 Bugzilla 965: Support mysql stored procedures.
2450 Patch from Alain Williams
2452 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2454 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2455 Patch from Andreas Metzler
2457 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2458 Patch from Kirill Miazine
2460 NM/05 Bugzilla 671: Added umask to procmail example.
2462 JJ/03 installed exipick 20100323.0, fixing doc bug
2464 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2465 directory. Notification and patch from Dan Rosenberg.
2467 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2469 TK/02 Improve log output when DKIM signing operation fails.
2471 MH/01 Treat the transport option dkim_domain as a colon separated
2472 list, not as a single string, and sign the message with each element,
2473 omitting multiple occurences of the same signer.
2475 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2476 Bugzilla 985, 986. Patch by Simon Arlott
2478 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2479 Patch by Simon Arlott
2481 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2482 MBX locking. Notification from Dan Rosenberg.
2488 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2490 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2492 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2494 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2496 NM/04 Bugzilla 915: Flag broken perl installation during build.
2502 TK/01 Added patch by Johannes Berg that expands the main option
2503 "spamd_address" if it starts with a dollar sign.
2505 TK/02 Write list of recipients to X-Envelope-Sender header when building
2506 the mbox-format spool file for content scanning (suggested by Jakob
2509 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2510 (http://www.dcc-servers.net/) support via dccifd. Activated by
2511 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2513 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2514 by Mark Daniel Reidel <mr@df.eu>.
2516 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2517 When building exim an external PCRE library is now needed -
2518 PCRE is a system library on the majority of modern systems.
2519 See entry on PCRE_LIBS in EDITME file.
2521 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2522 conversation. Added nologin parameter to request.
2523 Patch contributed by Kirill Miazine.
2525 TF/01 Do not log submission mode rewrites if they do not change the address.
2527 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2529 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2530 log files in place. Contributed by Roberto Lima.
2532 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2534 TF/03 Bugzilla 615: When checking the local_parts router precondition
2535 after a local_part_suffix or local_part_prefix option, Exim now
2536 does not use the address's named list lookup cache, since this
2537 contains cached lookups for the whole local part.
2539 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2540 Robert Millan. Documentation is in experimental-spec.txt.
2542 TF/04 Bugzilla 668: Fix parallel build (make -j).
2544 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2546 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2547 Patch provided by Jan Srzednicki.
2549 TF/05 Leading white space used to be stripped from $spam_report which
2550 wrecked the formatting. Now it is preserved.
2552 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2553 that they are available at delivery time.
2555 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2557 TF/08 TLS error reporting now respects the incoming_interface and
2558 incoming_port log selectors.
2560 TF/09 Produce a more useful error message if an SMTP transport's hosts
2561 setting expands to an empty string.
2563 NM/06 Bugzilla 744: EXPN did not work under TLS.
2564 Patch provided by Phil Pennock.
2566 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2567 Patch provided by Richard Godbee.
2569 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2570 acl_smtp_notquit, added index entry.
2572 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2573 Patch provided by Eugene Bujak.
2575 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2576 accept(). Patch provided by Maxim Dounin.
2578 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2579 Patch provided by Phil Pennock.
2581 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2583 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2584 Patch provided by Brad "anomie" Jorsch.
2586 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2587 Patch provided by Dean Brooks.
2589 TK/05 Add native DKIM support (does not depend on external libraries).
2591 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2592 Patch provided by Graeme Fowler.
2594 NM/16 Bugzilla 851: Documentation example syntax fix.
2596 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2598 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2601 NM/19 Bugzilla 745: TLS version reporting.
2602 Patch provided by Phil Pennock.
2604 NM/20 Bugzilla 167: bool: condition support.
2605 Patch provided by Phil Pennock.
2607 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2608 clients. Patch provided by Phil Pennock.
2610 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2611 Patch provided by Brad "anomie" Jorsch.
2613 NM/23 Bugzilla 687: Fix misparses in eximstats.
2614 Patch provided by Heiko Schlittermann.
2616 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2617 Patch provided by Heiko Schlittermann.
2619 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2620 plus update to original patch.
2622 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2624 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2625 Patch provided by David Brownlee.
2627 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2629 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2631 NM/30 Bugzilla 888: TLS documentation bugfixes.
2633 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2635 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2636 Unlike the original bugzilla I have changed all shell scripts in src tree.
2638 NM/33 Bugzilla 898: Transport filter timeout fix.
2639 Patch by Todd Rinaldo.
2641 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2642 Patch by Serge Demonchaux.
2644 NM/35 Bugzilla 39: Base64 decode bug fixes.
2645 Patch by Jakob Hirsch.
2647 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2649 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2651 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2653 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2659 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2660 ALT-N's libdkim that I have put here:
2661 http://duncanthrax.net/exim-experimental/
2663 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2664 were called 'true' and it seems that C99 defines that as a reserved
2665 keyword to be used with 'bool' variable types. That means you could
2666 not include C99-style headers which use bools without triggering
2667 build errors in sieve.c.
2669 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2670 as mailq or other aliases. Changed the --help handling significantly
2671 to do whats expected. exim_usage() emits usage/help information.
2673 SC/01 Added the -bylocaldomain option to eximstats.
2675 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2677 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2679 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2685 PH/01 Another patch from the Sieve maintainer.
2687 PH/02 When an IPv6 address is converted to a string for single-key lookup
2688 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2689 dots are used instead of colons so that keys in lsearch files need not
2690 contain colons. This was done some time before quoting was made available
2691 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2692 (notated using the quote facility) so as to distinguish them from IPv4
2693 keys. This meant that lookups for IP addresses in host lists did not work
2694 for iplsearch lookups.
2696 This has been fixed by arranging for IPv6 addresses to be expressed with
2697 colons if the lookup type is iplsearch. This is not incompatible, because
2698 previously such lookups could never work.
2700 The situation is now rather anomalous, since one *can* have colons in
2701 ordinary lsearch keys. However, making the change in all cases is
2702 incompatible and would probably break a number of configurations.
2704 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2707 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2708 conversion specification without a maximum field width, thereby enabling
2709 a rogue spamd server to cause a buffer overflow. While nobody in their
2710 right mind would setup Exim to query an untrusted spamd server, an
2711 attacker that gains access to a server running spamd could potentially
2712 exploit this vulnerability to run arbitrary code as the Exim user.
2714 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2715 $primary_hostname instead of what libspf2 thinks the hosts name is.
2717 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2718 a directory entry by the name of the lookup key. Previously, if a
2719 symlink pointed to a non-existing file or a file in a directory that
2720 Exim lacked permissions to read, a lookup for a key matching that
2721 symlink would fail. Now it is enough that a matching directory entry
2722 exists, symlink or not. (Bugzilla 503.)
2724 PH/03 The body_linecount and body_zerocount variables are now exported in the
2727 PH/04 Added the $dnslist_matched variable.
2729 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2730 This means they are set thereafter only if the connection becomes
2733 PH/06 Added the client_condition to authenticators so that some can be skipped
2734 by clients under certain conditions.
2736 PH/07 The error message for a badly-placed control=no_multiline_responses left
2737 "_responses" off the end of the name.
2739 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2741 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2742 (without spaces) instead of just copying the configuration text.
2744 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2746 PH/11 Added $max_received_linelength.
2748 PH/12 Added +ignore_defer and +include_defer to host lists.
2750 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2751 way in which PCRE > 7.0 is built.
2753 PH/14 Implemented queue_only_load_latch.
2755 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2756 MAIL command. The effect was to mangle the value on 64-bit systems.
2758 PH/16 Another patch from the Sieve maintainer.
2760 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2762 PH/18 If a system quota error occurred while trying to create the file for
2763 a maildir delivery, the message "Mailbox is full" was not appended to the
2764 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2765 only to a quota excession during the actual writing of the file.
2767 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2768 characters?) which causes problems in log lines. The DN values are now
2769 passed through string_printing() before being added to log lines.
2771 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2772 and InterBase are left for another time.)
2774 PH/21 Added message_body_newlines option.
2776 PH/22 Guard against possible overflow in moan_check_errorcopy().
2778 PH/23 POSIX allows open() to be a macro; guard against that.
2780 PH/24 If the recipient of an error message contained an @ in the local part
2781 (suitably quoted, of course), incorrect values were put in $domain and
2782 $local_part during the evaluation of errors_copy.
2788 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2789 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2792 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2793 issue a MAIL command.
2795 PH/02 In an ACL statement such as
2797 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2799 if a client was not listed at all, or was listed with a value other than
2800 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2801 the condition was not true (as it should be), so access was not denied.
2802 The bug was that the ! inversion was incorrectly passed on to the second
2803 item. This has been fixed.
2805 PH/03 Added additional dnslists conditions == and =& which are different from
2806 = and & when the dns lookup returns more than one IP address.
2808 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2809 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2811 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2812 FSYNC, which compiles an option called disable_fsync that allows for
2813 bypassing fsync(). The documentation is heavily laced with warnings.
2815 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2817 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2818 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2819 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2820 including adding "make clean"; (3) Added -fPIC when compiling the test
2821 dynamically loaded module, to get rid of a warning.
2823 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2824 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2825 = 0s. The bug is otherwise harmless.
2827 PH/07 There was a bug in the dovecot authenticator such that the value of
2828 $auth1 could be overwritten, and so not correctly preserved, after a
2829 successful authentication. This usually meant that the value preserved by
2830 the server_setid option was incorrect.
2832 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2834 PH/09 Installed PCRE release 7.0.
2836 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2837 run for batched SMTP input. It is now run at the start of every message
2838 in the batch. While fixing this I discovered that the process information
2839 (output by running exiwhat) was not always getting set for -bs and -bS
2840 input. This is fixed, and it now also says "batched" for BSMTP.
2842 PH/11 Added control=no_pipelining.
2844 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2845 patch, slightly modified), and move the expansion of helo_data till after
2846 the connection is made in the smtp transport (so it can use these
2849 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2851 PH/14 Added log_selector = +pid.
2853 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2855 PH/16 Add ${if forany and ${if forall.
2857 PH/17 Added dsn_from option to vary the From: line in DSNs.
2859 PH/18 Flush SMTP output before performing a callout, unless control =
2860 no_callout_flush is set.
2862 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2863 was true (the default) a successful delivery failed to delete the retry
2864 item, thus causing premature timeout of the address. The bug is now
2867 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2869 PH/21 Long custom messages for fakedefer and fakereject are now split up
2870 into multiline responses in the same way that messages for "deny" and
2871 other ACL rejections are.
2873 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2874 with slight modification.
2876 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2877 draft, changing the syntax and factoring some duplicate code".
2879 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2880 for deliveries of the second and subsequent messages over the same SMTP
2883 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2884 ${reduce, with only minor "tidies".
2886 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2888 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2889 expansion side effects.
2891 PH/27 When a message times out after an over-quota error from an Exim-imposed
2892 quota, the bounce message says "mailbox is full". This message was not
2893 being given when it was a system quota that was exceeded. It now should
2896 MH/03 Made $recipients available in local_scan(). local_scan() already has
2897 better access to the recipient list through recipients_list[], but
2898 $recipients can be useful in postmaster-provided expansion strings.
2900 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2901 in the case of a MAIL command with additional options following the
2902 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2903 were accidentally chopped off.
2905 PH/29 SMTP synchronization checks are implemented when a command is read -
2906 there is a check that no more input is waiting when there shouldn't be
2907 any. However, for some commands, a delay in an ACL can mean that it is
2908 some time before the response is written. In this time, more input might
2909 arrive, invalidly. So now there are extra checks after an ACL has run for
2910 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2911 pipelining has not been advertised.
2913 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2915 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2916 returned in the SMTP response when smtp_return_error_details was false.
2917 This has been fixed.
2919 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2920 instead of the C I/O that was originally supplied, because problems were
2921 reported on Solaris.
2923 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2924 Exim which did not show up earlier: it was assuming that a call to
2925 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2926 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2927 was a macro that became an assignment, so it seemed to work. This has
2928 changed to a proper function call with a void return, hence the compile
2929 error. Exim's code has been fixed.
2931 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2934 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2936 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2938 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2939 because it turns out to be too big or there is a timeout) there is no
2940 "Completed" line in the log. When some messages of this type were
2941 selected by exigrep, they were listed as "not completed". Others were
2942 picked up by some special patterns. I have improved the selection
2943 criteria to be more general.
2945 PH/38 The host_find_failed option in the manualroute router can now be set
2946 to "ignore", to completely ignore a host whose IP address cannot be
2947 found. If all hosts are ignored, the behaviour is controlled by the new
2948 host_all_ignored option.
2950 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2951 homing or because of multiple MX records with /mx) generated more than
2952 one IP address, and the following item turned out to be the local host,
2953 all the secondary addresses of the first item were incorrectly removed
2954 from the list, along with the local host and any following hosts (which
2955 is what is supposed to happen).
2957 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2958 whoever called Exim into the -H file. In the case of the daemon it was
2959 behaving confusingly. When first started, it used values for whoever
2960 started the daemon, but after a SIGHUP it used the Exim user (because it
2961 calls itself on a restart). I have changed the code so that it now always
2964 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2965 message are rejected with the same error (e.g. no authentication or bad
2966 sender address), and a DATA command is nevertheless sent (as can happen
2967 with PIPELINING or a stupid MUA), the error message that was given to the
2968 RCPT commands is included in the rejection of the DATA command. This is
2969 intended to be helpful for MUAs that show only the final error to their
2972 PH/42 Another patch from the Sieve maintainer.
2974 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2975 Eximstats - Fixed some broken HTML links and added missing column headers
2977 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2978 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2980 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2982 PH/43 Yet another patch from the Sieve maintainer.
2984 PH/44 I found a way to check for a TCP/IP connection going away before sending
2985 the response to the final '.' that terminates a message, but only in the
2986 case where the client has not sent further data following the '.'
2987 (unfortunately, this is allowed). However, in many cases there won't be
2988 any further data because there won't be any more messages to send. A call
2989 to select() can be used: if it shows that the input is "ready", there is
2990 either input waiting, or the socket has been closed. An attempt to read
2991 the next input character can distinguish the two cases. Previously, Exim
2992 would have sent an OK response which the client would never have see.
2993 This could lead to message repetition. This fix should cure that, at
2994 least in a lot of common cases.
2996 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2997 advertised in response to EHLO.
3003 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3004 fixed by 4.65/MH/01 (is this a record?) are fixed:
3006 (i) An empty string was always treated as zero by the numeric comparison
3007 operators. This behaviour has been restored.
3009 (ii) It is documented that the numeric comparison operators always treat
3010 their arguments as decimal numbers. This was broken in that numbers
3011 starting with 0 were being interpreted as octal.
3013 While fixing these problems I realized that there was another issue that
3014 hadn't been noticed. Values of message_size_limit (both the global option
3015 and the transport option) were treated as octal if they started with 0.
3016 The documentation was vague. These values are now always treated as
3017 decimal, and I will make that clear in the documentation.
3023 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3024 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3027 MH/01 Don't check that the operands of numeric comparison operators are
3028 integers when their expansion is in "skipping" mode (fixes bug
3029 introduced by 4.64-PH/07).
3031 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3032 child addresses, Exim now panics and dies. Previously, because the count
3033 is held in a short int, deliveries were likely to be lost. As such a
3034 large number of recipients for a single message is ridiculous
3035 (performance will be very, very poor), I have chosen to impose a limit
3036 rather than extend the field.
3042 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3043 leftover -K file (the existence of which was triggered by #402).
3044 While we were at it, introduced process PID as part of the -K
3045 filename. This should rule out race conditions when creating
3048 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3049 processing considerably. Previous code took too long for large mails,
3050 triggering a timeout which in turn triggers #401.
3052 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3053 in the DK code in transports.c. sendfile() is not really portable,
3054 hence the _LINUX specificness.
3056 TF/01 In the add_headers option to the mail command in an Exim filter,
3057 there was a bug that Exim would claim a syntax error in any
3058 header after the first one which had an odd number of characters
3061 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3062 callout verification, Exim cached a "reject" for the entire domain. This
3063 is correct for most verifications, but it is not correct for a recipient
3064 verification with use_sender or use_postmaster set, because in that case
3065 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3066 case of MAIL FROM:<> rejection from other early rejections (e.g.
3067 rejection of HELO). When verifying a recipient using a non-null MAIL
3068 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3069 Whatever the result of the callout, the value of the domain cache is
3070 left unchanged (for any other kind of callout, getting as far as trying
3071 RCPT means that the domain itself is ok).
3073 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3076 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3077 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3078 session, the daemon ignores ECONNECT errors and logs others; it now
3079 ignores EPIPE as well.
3081 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3082 (quoted-printable decoding).
3084 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3085 later the small subsequent patch to fix an introduced bug.
3087 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3089 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3091 PH/08 An error is now given if message_size_limit is specified negative.
3093 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3094 to be given (somewhat) arbitrary names.
3096 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3099 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3100 miscellaneous code fixes
3102 PH/10 Added the log_reject_target ACL modifier to specify where to log
3105 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3106 hostname. This is wrong, because it relates to the incoming message (and
3107 probably the interface on which it is arriving) and not to the outgoing
3108 callout (which could be using a different interface). This has been
3109 changed to use the value of the helo_data option from the smtp transport
3110 instead - this is what is used when a message is actually being sent. If
3111 there is no remote transport (possible with a router that sets up host
3112 addresses), $smtp_active_hostname is used.
3114 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3115 tweaks were necessary in order to get it to work (see also 21 below):
3116 (a) The code assumed that strncpy() returns a negative number on buffer
3117 overflow, which isn't the case. Replaced with Exim's string_format()
3119 (b) There were several signed/unsigned issues. I just did the minimum
3120 hacking in of casts. There is scope for a larger refactoring.
3121 (c) The code used strcasecmp() which is not a standard C function.
3122 Replaced with Exim's strcmpic() function.
3123 (d) The code set only $1; it now sets $auth1 as well.
3124 (e) A simple test gave the error "authentication client didn't specify
3125 service in request". It would seem that Dovecot has changed its
3126 interface. Fortunately there's a specification; I followed it and
3127 changed what the client sends and it appears to be working now.
3129 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3132 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3133 address A is aliased to B and C, where B exists and C does not. Without
3134 -v the output is "A verified" because verification stops after a
3135 successful redirection if more than one address is generated. However,
3136 with -v the child addresses are also verified. Exim was outputting "A
3137 failed to verify" and then showing the successful verification for C,
3138 with its parentage. It now outputs "B failed to verify", showing B's
3139 parentage before showing the successful verification of C.
3141 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3142 look up a TXT record in a specific list after matching in a combined
3145 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3146 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3147 they consult the DNS. I had assumed they would set it the way they
3148 wanted; and indeed my experiments on Linux seem to show that in some
3149 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3150 To be on the safe side, however, I have now made the interface to
3151 host_find_byname() similar to host_find_bydns(), with an argument
3152 containing the DNS resolver options. The host_find_byname() function now
3153 sets these options at its start, just as host_find_bydns() does. The smtp
3154 transport options dns_qualify_single and dns_search_parents are passed to
3155 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3156 of host_find_byname() use the default settings of RES_DEFNAMES
3157 (qualify_single) but not RES_DNSRCH (search_parents).
3159 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3160 spool_read_header() do less string testing, by means of a preliminary
3161 switch on the second character of optional "-foo" lines. (This is
3162 overdue, caused by the large number of possibilities that now exist.
3163 Originally there were few.) While I was there, I also converted the
3164 str(n)cmp tests so they don't re-test the leading "-" and the first
3165 character, in the hope this might squeeze out yet more improvement.
3167 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3168 flag allowing group syntax was set by the header_syntax check but not
3169 turned off, possible causing trouble later; (2) The flag was not being
3170 set at all for the header_verify test, causing "group"-style headers to
3171 be rejected. I have now set it in this case, and also caused header_
3172 verify to ignore an empty address taken from a group. While doing this, I
3173 came across some other cases where the code for allowing group syntax
3174 while scanning a header line wasn't quite right (mostly, not resetting
3175 the flag correctly in the right place). These bugs could have caused
3176 trouble for malformed header lines. I hope it is now all correct.
3178 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3179 with the "reply" argument non-NULL. The code, however (which originally
3180 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3181 but it didn't always do it. This confused somebody who was copying the
3182 code for some other use. I have removed all the tests.
3184 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3185 feature that was used to support insecure browsers during the U.S. crypto
3186 embargo. It requires special client support, and Exim is probably the
3187 only MTA that supported it -- and would never use it because real RSA is
3188 always available. This code has been removed, because it had the bad
3189 effect of slowing Exim down by computing (never used) parameters for the
3190 RSA_EXPORT functionality.
3192 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3193 authenticator to fail if there's a tab character in the incoming data
3194 (there should never be unless someone is messing about, as it's supposed
3195 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3196 if the connection is using TLS or if the remote IP is the same as the
3197 local IP, and the "valid-client-cert option" if a client certificate has
3200 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3201 authenticators. This can be used for authorization after authentication
3202 succeeds. (In the case of plaintext, it servers for both authentication
3205 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3206 if any retry times were supplied.
3208 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3209 connection, where there is no client IP address to check. In this
3210 situation, the verify now always succeeds.
3212 PH/25 Applied John Jetmore's -Mset patch.
3214 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3216 PH/27 In a string expansion for a processed (not raw) header when multiple
3217 headers of the same name were present, leading whitespace was being
3218 removed from all of them, but trailing whitespace was being removed only
3219 from the last one. Now trailing whitespace is removed from each header
3220 before concatenation. Completely empty headers in a concatenation (as
3221 before) are ignored.
3223 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3224 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3226 PH/29 [Removed. This was a change that I later backed out, and forgot to
3227 correct the ChangeLog entry (that I had efficiently created) before
3228 committing the later change.]
3230 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3231 address errors (4xx response to RCPT) over the same connection as other
3232 messages routed to the same hosts. Such deliveries are always "forced",
3233 so retry times are not inspected. This resulted in far too many retries
3234 for the affected addresses. The effect occurred only when there were more
3235 hosts than the hosts_max_try setting in the smtp transport when it had
3236 the 4xx errors. Those hosts that it had tried were not added to the list
3237 of hosts for which the message was waiting, so if all were tried, there
3238 was no problem. Two fixes have been applied:
3240 (i) If there are any address or message errors in an SMTP delivery, none
3241 of the hosts (tried or untried) are now added to the list of hosts
3242 for which the message is waiting, so the message should not be a
3243 candidate for sending over the same connection that was used for a
3244 successful delivery of some other message. This seems entirely
3245 reasonable: after all the message is NOT "waiting for some host".
3246 This is so "obvious" that I'm not sure why it wasn't done
3247 previously. Hope I haven't missed anything, but it can't do any
3248 harm, as the worst effect is to miss an optimization.
3250 (ii) If, despite (i), such a delivery is accidentally attempted, the
3251 routing retry time is respected, so at least it doesn't keep
3252 hammering the server.
3254 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3255 in ${readsocket because some servers need this prod.
3257 PH/32 Added some extra debug output when updating a wait-xxx database.
3259 PH/33 The hint "could be header name not terminated by colon", which has been
3260 given for certain expansion errors for a long time, was not being given
3261 for the ${if def:h_colon_omitted{... case.
3263 PH/34 The spec says: "With one important exception, whenever a domain list is
3264 being scanned, $domain contains the subject domain." There was at least
3265 one case where this was not true.
3267 PH/35 The error "getsockname() failed: connection reset by peer" was being
3268 written to the panic log as well as the main log, but it isn't really
3269 panic-worthy as it just means the connection died rather early on. I have
3270 removed the panic log writing for the ECONNRESET error when getsockname()
3273 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3274 runs only) independently of the message's sender address. This meant
3275 that, if the 4xx error was in fact related to the sender, a different
3276 message to the same recipient with a different sender could confuse
3277 things. In particular, this can happen when sending to a greylisting
3278 server, but other circumstances could also provoke similar problems.
3279 I have changed the default so that the retry time for these errors is now
3280 based a combination of the sender and recipient addresses. This change
3281 can be overridden by setting address_retry_include_sender=false in the
3284 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3285 remote server are returned as part of bounce messages. This was not
3286 happening for LMTP over a pipe (the lmtp transport), but now it is the
3287 same for both kinds of LMTP.
3289 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3290 in header lines that were in fact CNAMEs. This is no longer the case.
3292 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3293 and queue runs started by the daemon processed all messages. This has
3294 been fixed so that -R and -S can now usefully be given with -q<time>.
3296 PH/40 Import PCRE release 6.7 (fixes some bugs).
3298 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3300 PH/42 Give an error if -q is specified more than once.
3302 PH/43 Renamed the variables $interface_address and $interface_port as
3303 $received_ip_address and $received_port, to make it clear that these
3304 values apply to message reception, and not to the outgoing interface when
3305 a message is delivered. (The old names remain recognized, of course.)
3307 PH/44 There was no timeout on the connect() call when using a Unix domain
3308 socket in the ${readsocket expansion. There now is.
3310 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3311 be meaningful with "accept".
3313 SC/01 Eximstats V1.43
3314 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3316 SC/02 Eximstats V1.44
3317 Use a glob alias rather than an array ref in the generated
3318 parser. This improves both readability and performance.
3320 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3321 Collect SpamAssassin and rejection statistics.
3322 Don't display local sender or destination tables unless
3323 there is data to show.
3324 Added average volumes into the top table text output.
3326 SC/04 Eximstats V1.46
3327 Collect data on the number of addresses (recipients)
3328 as well as the number of messages.
3330 SC/05 Eximstats V1.47
3331 Added 'Message too big' to the list of mail rejection
3332 reasons (thanks to Marco Gaiarin).
3334 SC/06 Eximstats V1.48
3335 Mainlog lines which have GMT offsets and are too short to
3336 have a flag are now skipped.
3338 SC/07 Eximstats V1.49 (Alain Williams)
3339 Added the -emptyok flag.
3341 SC/08 Eximstats V1.50
3342 Fixes for obtaining the IP address from reject messages.
3344 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3345 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3346 whitespace changes from 4.64-PH/27
3348 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3351 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3352 are found, allow negative numbers in numeric criteria)
3354 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3356 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3357 to match changes made in 4.64-PH/43
3359 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3361 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3362 "domain" is actually an IP address, and force a failure. This locks out
3363 those revolvers/nameservers that support "A-for-A" lookups, in
3364 contravention of the specifications.
3366 PH/48 When a host name was looked up from an IP address, and the subsequent
3367 forward lookup of the name timed out, the host name was left in
3368 $sender_host_name, contrary to the specification.
3370 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3371 restricted to single-key lookups, Exim was not diagnosing an error if
3372 * or *@ was used with a query-style lookup.
3374 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3376 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3377 long ago, but noone interested enough thought of it. Let's just say that
3378 the "1.1" means that there are some new functions that weren't there at
3379 some point in the past.
3381 PH/51 Error processing for expansion failure of helo_data from an smtp
3382 transport during callout processing was broken.
3384 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3385 tested/used via the -bh/-bhc/-bs options.
3387 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3388 bug, fixed in subsequent PCRE releases).
3390 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3391 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3393 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3399 SC/01 Use a glob alias rather than an array ref in eximstats generated
3400 parser. This improves both readability and performance.
3402 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3403 Don't display local sender or destination tables in eximstats unless
3404 there is data to show.
3405 Added average volumes into the eximstats top table text output.
3407 SC/03 Collect data on the number of addresses (recipients) as well
3408 as the number of messages in eximstats.
3410 TF/01 Correct an error in the documentation for the redirect router. Exim
3411 does (usually) call initgroups() when daemonizing.
3413 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3414 with consistent privilege compared to when running as a daemon.
3416 TF/03 Note in the spec that $authenticated_id is not set for local
3417 submissions from trusted users.
3419 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3420 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3422 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3423 by adding some example configuration directives to the default
3424 configuration file. A little bit of work is required to uncomment the
3425 directives and define how usernames and passwords are checked, but
3426 there is now a framework to start from.
3428 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3429 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3430 without this. I don't know how relevant this is to other LDAP libraries.
3432 PH/02 Add the verb name to the "unknown ACL verb" error.
3434 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3436 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3438 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3439 directory not expanded when it should be if an expanded home directory
3440 was set for the address (which is overridden by the transport).
3442 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3445 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3446 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3447 because it is too late at that time, and has no effect.
3449 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3450 security issue with \' (bugzilla #107). I could not use the
3451 PQescapeStringConn() function, because it needs a PGconn value as one of
3454 PH/08 When testing addresses using -bt, indicate those final addresses that
3455 are duplicates that would not cause an additional delivery. At least one
3456 person was confused, thinking that -bt output corresponded to deliveries.
3457 (Suppressing duplicates isn't a good idea as you lose the information
3458 about possibly different redirections that led to the duplicates.)
3460 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3461 systems where poll() doesn't work, in particular OS X.
3463 PH/10 Added more information to debugging output for retry time not reached.
3465 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3466 operations in malware.c.
3468 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3471 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3472 syslog_duplication set false, log lines that would normally be written
3473 both the the main log and to the reject log were not written to syslog at
3476 PH/14 In the default configuration, change the use of "message" in ACL warn
3477 statements to "add_header".
3479 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3480 not followed by a command (e.g. "seen endif").
3482 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3483 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3486 PH/17 Added extra conditions to the default value of delay_warning_condition
3490 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3491 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3492 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3495 The Auto-Submitted: and various List- headers are standardised, whereas I
3496 don't think Precedence: ever was.
3498 PH/18 Refactored debugging code in route_finduser() to show more information,
3499 in particular, the error code if getpwnam() issues one.
3501 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3502 This is apparently needed in addition to the PH/07 change above to avoid
3503 any possible encoding problems.
3505 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3506 but not after initializing Perl.
3508 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3509 output them only if debugging. By default they are written stderr,
3510 apparently, which is not desirable.
3512 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3515 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3518 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3520 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3521 authenticated or an ident call has been made. Suppress the default
3522 values for $authenticated_id and $authenticated_sender (but permit -oMai
3523 and -oMas) when testing with -bh.
3525 PH/24 Re-jigged the order of the tests in the default configuration so that the
3526 tests for valid domains and recipients precede the DNS black list and CSA
3527 tests, on the grounds that those ones are more expensive.
3529 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3530 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3531 This bug exists in every version of Exim that I still have, right back to
3534 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3535 However, an attempt to turn on case-sensitivity in a regex key by
3536 including (?-i) didn't work because the subject string was already
3537 lowercased, and the effects were non-intuitive. It turns out that a
3538 one-line patch can be used to allow (?-i) to work as expected.
3544 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3545 other effects) broke the use of negated acl sub-conditions.
3547 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3550 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3551 "Deny" causes Exim to reject the incoming connection with a 554 error.
3552 Unfortunately, if there is a major crisis, such as a disk failure,
3553 tcp-wrappers gives "deny", whereas what one would like would be some
3554 kind of temporary error. A kludge has been added to help with this.
3555 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3556 554 error is used if errno is still zero or contains ENOENT (which occurs
3557 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3560 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3562 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3563 errors. Otherwise a message that provokes a temporary error (when other
3564 messages do not) can cause a whole host to time out.
3566 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3567 addresses were routed directly to files or pipes from a redirect router.
3568 File deliveries just didn't batch; pipe deliveries might have suffered
3571 PH/06 A failure to get a lock for a hints database would erroneously always say
3572 "Failed to get write lock", even when it was really a read lock.
3574 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3575 of 0600. This has been changed to use the value of the lockfile_mode
3576 option (which defaults to 0600).
3578 PH/08 Applied small patch from the Sieve maintainer.
3580 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3581 folder from quota calculations, a direct delivery into this folder messed
3582 up the contents of the maildirsize file. This was because the regex was
3583 used only to exclude .Trash (or whatever) when the size of the mailbox
3584 was calculated. There was no check that a delivery was happening into an
3585 excluded directory. This bug has been fixed by ignoring all quota
3586 processing for deliveries into excluded directories.
3588 PH/10 Added the maildirfolder_create_regex option to appendfile.
3594 PH/01 The code for finding all the local interface addresses on a FreeBSD
3595 system running IPv6 was broken. This may well have applied to all BSD
3596 systems, as well as to others that have similar system calls. The broken
3597 code found IPv4 interfaces correctly, but gave incorrect values for the
3598 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3599 that it would not match correctly against @[] and not recognize the IPv6
3602 PH/02 The ipliteral router was not recognizing addresses of the form user@
3603 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3605 PH/03 Added disable_ipv6.
3607 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3608 decoded form, because it is most often used to construct To: headers
3609 lines in autoreplies, and the decoded form may well be syntactically
3610 invalid. However, $reply_address has leading white space removed, and all
3611 newlines turned into spaces so that the autoreply transport does not
3614 PH/05 If group was specified without a user on a router, and no group or user
3615 was specified on a transport, the group from the router was ignored.
3617 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3618 for visible compile-time settings that can be used to change these
3619 numbers, for those that want even more. Backwards compatibility with old
3620 spool files has been maintained. However, going back to a previous Exim
3621 release will lost any variables that are in spool files.
3623 PH/07 Two small changes when running in the test harness: increase delay when
3624 passing a TCP/IP connection to a new process, in case the original
3625 process has to generate a bounce, and remove special handling of
3626 127.0.0.2 (sic), which is no longer necessary.
3628 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3629 be the same on different OS.
3631 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3634 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3635 whether --show-vars was specified or not
3637 JJ/02 exipick: Added support for new ACL variable spool format introduced
3640 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3641 syntactically invalid From: or Reply-to: line, and a filter used this to
3642 generate an autoreply, and therefore failed to obtain an address for the
3643 autoreply, Exim could try to deliver to a non-existent relative file
3644 name, causing unrelated and misleading errors. What now happens is that
3645 it logs this as a hard delivery error, but does not attempt to create a
3648 PH/11 The exinext utility has a -C option for testing purposes, but although
3649 the given file was scanned by exinext itself; it wasn't being passed on
3650 when Exim was called.
3652 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3653 an end-of-file indication when reading a command response.
3655 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3656 compiled. In many other places in Exim, IPv6 addresses are always
3657 recognized, so I have changed this. It also means that IPv4 domain
3658 literals of the form [IPV4:n.n.n.n] are now always recognized.
3660 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3661 used if the router is not running as root, for example, when verifying at
3662 ACL time, or when using -bh. The debugging output from this situation was
3663 non-existent - all you got was a failure to exec. I have made two
3666 (a) Failures to set uid/gid, the current directory, or a process leader
3667 in a subprocess such as that created by queryprogram now generate
3668 suitable debugging output when -d is set.
3670 (b) The queryprogram router detects when it is not running as root,
3671 outputs suitable debugging information if -d is set, and then runs
3672 the subprocess without attempting to change uid/gid.
3674 PH/15 Minor change to Makefile for building test_host (undocumented testing
3677 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3678 additional section of a DNS packet that returns MX or SRV records.
3679 Instead, it always explicitly searches for A/AAAA records. This avoids
3680 major problems that occur when a DNS server includes only records of one
3681 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3682 fixed another bug: if SRV records were looked up and the corresponding
3683 address records were *not* found in the additional section, the port
3684 values from the SRV records were lost.
3686 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3687 using the correct key (the original address) when searching the retry
3688 rules in order to find which one to use for generating the retry hint.
3690 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3691 adding the default one. Similarly, if it contains a Reply-To: header, the
3692 errors_reply_to option, if set, is not used.
3694 PH/19 When calculating a retry time, Exim used to measure the "time since
3695 failure" by looking at the "first failed" field in the retry record. Now
3696 it does not use this if it is later than than the arrival time of the
3697 message. Instead it uses the arrival time. This makes for better
3698 behaviour in cases where some deliveries succeed, thus re-setting the
3699 "first failed" field. An example is a quota failure for a huge message
3700 when small messages continue to be delivered. Without this change, the
3701 "time since failure" will always be short, possible causing more frequent
3702 delivery attempts for the huge message than are intended.
3703 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3705 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3706 $1, $2, $3) because the numerical variables can be reset during some
3707 expansion items (e.g. "match"), thereby losing the authentication data.
3709 PH/21 Make -bV show the size of off_t variables so that the test suite can
3710 decide whether to run tests for quotas > 2G.
3712 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3713 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3714 filecount value is greater than 2G or if a quota value is greater than 2G
3715 on a system where the size of off_t is not greater than 4, a panic error
3718 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3719 never match. The debug and -bh output now contains an explicit error
3720 message indicating a malformed IPv4 address or mask.
3722 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3723 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3724 PH/23 above applies.
3726 PH/25 Do not write to syslog when running in the test harness. The only
3727 occasion when this arises is a failure to open the main or panic logs
3728 (for which there is an explicit test).
3730 PH/26 Added the /no_tell option to "control=freeze".
3732 PH/27 If a host name lookup failed very early in a connection, for example, if
3733 the IP address matched host_lookup and the reverse lookup yielded a name
3734 that did not have a forward lookup, an error message of the form "no IP
3735 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3736 could be logged. Now it outputs the IP address instead of "NULL".
3738 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3739 allows the sender and the authenticated sender to be set when
3740 submitting a message from within Exim. Since child_open_exim() is
3741 documented for local_scan(), the new function should be too.
3743 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3744 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3745 results in an empty string is now treated as unset.
3747 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3749 PH/31 Added sender_verify_fail logging option.
3751 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3752 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3753 tidied the source and removed it altogether.
3755 PH/33 When a queue run was abandoned because the load average was too high, a
3756 log line was always written; now it is written only if the queue_run log
3757 selector is set. In addition, the log line for abandonment now contains
3758 information about the queue run such as the pid. This is always present
3759 in "start" and "stop" lines but was omitted from the "abandon" line.
3761 PH/34 Omit spaces between a header name and the colon in the error message that
3762 is given when verify = headers_syntax fails (if there are lots of them,
3763 the message gets confusing).
3765 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3766 names, as there are now some PTR records that contain slashes. This check
3767 is only to protect against broken name servers that fall over on strange
3768 characters, so the fact that it applies to all lookups doesn't matter.
3770 PH/36 Now that the new test suite is complete, we can remove some of the
3771 special code in Exim that was needed for the old test suite. For example,
3772 sorting DNS records because real resolvers return them in an arbitrary
3773 order. The new test suite's fake resolver always returns records in the
3776 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3777 bounces) except when queue_only is set, to avoid logging races between
3778 the different processes.
3780 PH/38 Panic-die if .include specifies a non-absolute path.
3782 PH/39 A tweak to the "H" retry rule from its user.
3784 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3785 a label. They prevented compilation on older perls.
3787 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3788 a warning to be raised on newish perls.
3790 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3791 on queue. Changes to match documented behaviour of showing count of
3792 messages matching specified criteria.
3794 PH/40 Changed the default ident timeout from 30s to 5s.
3796 PH/41 Added support for the use of login_cap features, on those BSD systems
3797 that have them, for controlling the resources used by pipe deliveries.
3799 PH/42 The content-scanning code uses fopen() to create files in which to put
3800 message data. Previously it was not paying any attention to the mode of
3801 the files. Exim runs with umask(0) because the rest of the code creates
3802 files with open(), and sets the required mode explicitly. Thus, these
3803 files were ending up world-writeable. This was not a big issue, because,
3804 being within the spool directory, they were not world-accessible. I have
3805 created a function called modefopen, which takes an additional mode
3806 argument. It sets umask(777), creates the file, chmods it to the required
3807 mode, then resets the umask. All the relevant calls to fopen() in the
3808 content scanning code have been changed to use this function.
3810 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3811 to 24 hours. This avoids potential overflow problems when processing G
3812 and H retry rules. I suspect nobody ever tinkers with this value.
3814 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3816 PH/45 When the plaintext authenticator is running as a client, the server's
3817 challenges are checked to ensure they are valid base64 strings. By
3818 default, the authentication attempt is cancelled if an invalid string is
3819 received. Setting client_ignore_invalid_base64 true ignores these errors.
3820 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3821 they are received. Thus, the responses can be made to depend on the
3822 challenges. If an invalid string is ignored, an empty string is placed in
3825 PH/46 Messages that are created by the autoreply transport now contains a
3826 References: header, in accordance with RFCs 2822 and 3834.
3828 PH/47 Added authenticated_sender_force to the smtp transport.
3830 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3832 PH/49 Installed latest patch from the Sieve maintainer.
3834 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3835 was also set, the appendfile transport was unnecessarily scanning a
3836 directory of message files (e.g. for maildir delivery) to find the count
3837 of files (along with the size), even though it did not need this
3838 information. It now does the scan only if it needs to find either the
3839 size of the count of files.
3841 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3843 PH/52 Two bugs concerned with error handling when the smtp transport is
3846 (i) Exim was not creating retry information for temporary errors given
3847 for individual recipients after the DATA command when the smtp transport
3848 was used in LMTP mode. This meant that they could be retried too
3849 frequently, and not timed out correctly.
3851 (ii) Exim was setting the flag that allows error details to be returned
3852 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3853 recipients that were returned after the DATA command.
3855 PH/53 This is related to PH/52, but is more general: for any failing address,
3856 when detailed error information was permitted to be returned to the
3857 sender, but the error was temporary, then after the final timeout, only
3858 "retry timeout exceeded" was returned. Now it returns the full error as
3859 well as "retry timeout exceeded".
3861 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3862 do this, and (what is worse) MTAs that accept it.
3864 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3865 will now be deprecated.
3867 PH/56 New os.c-cygwin from the Cygwin maintainer.
3869 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3870 formats (previously only available in exim formats via -bpr, -bpru,
3871 and -bpra. Now also available in native and exiqgrep formats)
3873 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3874 with very large, slow to parse queues
3876 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3878 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3880 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3881 responses to authentication challenges, though it was showing the
3882 challenges; (ii) I've removed the CR characters from the debug output for
3885 PH/58 Allow for the insertion of a newline as well as a space when a string
3886 is turned into more than one encoded-word during RFC 2047 encoding. The
3887 Sieve code now uses this.
3889 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3890 data_4xx, lost_connection, tls_required.
3892 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3893 message was being sent as an SMTP response.
3895 PH/61 Add -l and -k options to exicyclog.
3897 PH/62 When verifying, if an address was redirected to one new address, so that
3898 verification continued, and the new address failed or deferred after
3899 having set something in $address_data, the value of $address_data was not
3900 passed back to the ACL. This was different to the case when no
3901 redirection occurred. The value is now passed back in both cases.
3903 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3904 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3905 use login_cap.h, so on its own it isn't the distinguishing feature. The
3906 new name refers directly to the setclassresources() function.
3908 PH/65 Added configuration files for NetBSD3.
3910 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3912 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3913 is preferred over IPv4.
3915 PH/68 The bounce_return_message and bounce_return_body options were not being
3916 honoured for bounces generated during the reception of non-SMTP messages.
3917 In particular, this applied to messages rejected by the ACL. This bug has
3918 been fixed. However, if bounce_return_message is true and bounce_return_
3919 body is false, the headers that are returned for a non-SMTP message
3920 include only those that have been read before the error was detected.
3921 (In the case of an ACL rejection, they have all been read.)
3923 PH/69 The HTML version of the specification is now built in a directory called
3924 spec_html instead of spec.html, because the latter looks like a path with
3925 a MIME-type, and this confuses some software.
3927 PH/70 Catch two compiler warnings in sieve.c.
3929 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3930 function verify_get_ident() calls ip_connect() to connect a socket, but
3931 if the "connect()" function timed out, ip_connect() used to close the
3932 socket. However, verify_get_ident() also closes the socket later, and in
3933 between Exim writes to the log, which may get opened at this point. When
3934 the socket was closed in ip_connect(), the log could get the same file
3935 descriptor number as the socket. This naturally causes chaos. The fix is
3936 not to close the socket in ip_connect(); the socket should be closed by
3937 the function that creates it. There was only one place in the code where
3938 this was missing, in the iplookup router, which I don't think anybody now
3939 uses, but I've fixed it anyway.
3941 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3942 well as to direct DNS lookups. Otherwise the handling of names in host
3943 lists is inconsistent and therefore confusing.
3949 PH/01 Two changes to the default runtime configuration:
3951 (1) Move the checks for relay_from_hosts and authenticated clients from
3952 after to before the (commented out) DNS black list checks.
3954 (2) Add control=submission to the relay_from_hosts and authenticated
3955 clients checks, on the grounds that messages accepted by these
3956 statements are most likely to be submissions.
3958 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3960 (1) Generate an error if the third argument for the ${prvs expansion is
3963 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3966 (3) Reset the variables that are obtained from the first argument of
3967 ${prvscheck and used in the second argument before leaving the code,
3968 because their memory is reclaimed, so using them afterwards may do
3971 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3972 one (it's much easier than Tom thought :-).
3974 (5) Because of (4), we can now allow for the use of $prvscheck_result
3975 inside the third argument.
3977 PH/03 For some reason, the default setting of PATH when running a command from
3978 a pipe transport was just "/usr/bin". I have changed it to
3981 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3982 anything to be listed in the output from -bV.
3984 PH/05 When a filter generated an autoreply, the entire To: header line was
3985 quoted in the delivery log line, like this:
3987 => >A.N.Other <ano@some.domain> <original@ddress> ...
3989 This has been changed so that it extracts the operative address. There
3990 may be more than one such address. If so, they are comma-separated, like
3993 => >ano@some.domain,ona@other.domain <original@ddress> ...
3995 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3996 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3997 not being looked up in the rDNS to get a host name, Exim was showing the
3998 IP address twice in Received: lines, even though the IP addresses were
3999 identical. For example:
4001 Received: from [1.2.3.4] (helo=[1.2.3.4])
4003 However, if the real host name was known, it was omitting the HELO data
4004 if it matched the actual IP address. This has been tidied up so that it
4005 doesn't show the same IP address twice.
4007 PH/07 When both +timestamp and +memory debugging was on, the value given by
4008 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4009 called by the debug printing, thereby overwriting the timestamp buffer.
4010 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4012 PH/08 When the original message was included in an autoreply transport, it
4013 always said "this is a copy of the message, including all the headers",
4014 even if body_only or headers_only was set. It now gives an appropriate
4017 PH/09 Applied a patch from the Sieve maintainer which:
4019 o fixes some comments
4020 o adds the (disabled) notify extension core
4021 o adds some debug output for the result of if/elsif tests
4022 o points to the current vacation draft in the documentation
4023 and documents the missing references header update
4027 o fixes a bug in processing the envelope test (when testing
4028 multiple envelope elements, the last element determined the
4031 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4032 Electronic Mail") by including:
4034 Auto-submitted: auto-generated
4036 in the messages that it generates (bounce messages and others, such as
4037 warnings). In the case of bounce messages for non-SMTP messages, there was
4038 also a typo: it was using "Auto_submitted" (underscore instead of
4039 hyphen). Since every message generated by Exim is necessarily in response
4040 to another message, thes have all been changed to:
4042 Auto-Submitted: auto-replied
4044 in accordance with these statements in the RFC:
4046 The auto-replied keyword:
4048 - SHOULD be used on messages sent in direct response to another
4049 message by an automatic process,
4051 - MUST NOT be used on manually-generated messages,
4053 - MAY be used on Delivery Status Notifications (DSNs) and Message
4054 Disposition Notifications (MDNs),
4056 - MUST NOT be used on messages generated by automatic or periodic
4057 processes, except for messages which are automatic responses to
4060 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4061 to the default Received: header definition.
4063 PH/12 Added log selector acl_warn_skipped (default on).
4065 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4066 variables because (a) they are in the wrong storage pool and (b) even if
4067 they were copied, it wouldn't work properly because of the caching.
4069 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4070 checking when decoding. Apparently there are clients that generate
4071 overlong encoded strings. Why am I not surprised?
4073 PH/15 If the first argument of "${if match_address" was not empty, but did not
4074 contain an "@" character, Exim crashed. Now it writes a panic log message
4075 and treats the condition as false.
4077 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4079 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4080 extension "envelope-auth". The code is finished and in agreement with
4081 other implementations, but there is no documentation so far and in fact,
4082 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4083 not changing the active code.
4085 Print executed "if" and "elsif" statements when debugging is used. This
4086 helps a great deal to understand what a filter does.
4088 Document more things not specified clearly in RFC3028. I had all this
4089 sorted out, when out of a sudden new issues came to my mind. Oops."
4091 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4094 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4095 canonical form (as documented). However, after a host name lookup from
4096 the IP address, check_host() was doing a simple string comparison with
4097 addresses acquired from the DNS when checking that the found name did
4098 have the original IP as one of its addresses. Since any found IPv6
4099 addresses are likely to be in abbreviated form, the comparison could
4100 fail. Luckily, there already exists a function for doing the comparison
4101 by converting both addresses to binary, so now that is used instead of
4102 the text comparison.
4104 PH/20 There was another similar case to PH/19, when a complete host name was
4105 given in a host list; looking up its IP address could give an abbreviated
4106 form, whereas the current host's name might or might not be abbreviated.
4107 The same fix has been applied.
4113 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4114 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4117 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4118 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4120 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4122 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4123 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4124 Also, for "Auto-Submitted", treat anything other than "no" as
4125 non-personal, in accordance with RFC 3834. (Previously it treated
4126 anything starting "auto-" as non-personal.)
4128 TF/01 The control=submission/name=... option had a problem with syntax
4129 errors if the name included a slash character. The /name= option
4130 now slurps the rest of the string, so it can include any characters
4131 but it must come last in the list of options (after /sender_retain
4134 PH/05 Some modifications to the interface to the fake nameserver for the new
4142 TK/01 Added the "success_on_redirect" address verification option. See
4143 NewStuff for rationale and an example.
4145 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4147 PH/02 Patch to exigrep to allow it to work on syslog lines.
4149 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4150 fread() to skip over the body file's header line, because in Cygwin the
4151 header line is locked and is inaccessible.
4153 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4154 co-exist for some time) to make it clear that it is the Exim ID that is
4155 referenced, not the Message-ID: header line.
4157 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4158 string_format() function, because snprintf() does not exist on all
4161 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4162 ${stat: expansion item.
4164 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4165 protocol synchronization error", to keep the pedants happy.
4167 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4168 well as for IRIX systems, when gcc is being used. See the host.c source
4171 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4173 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4176 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4177 transport and to the smtp transport in LMTP mode.
4179 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4181 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4182 run a filter in a subprocess. This could lead to confusion in subsequent
4183 lookups in the parent process. There should also be a search_tidyup() at
4184 the end of the subprocess.
4186 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4187 only if the host matched helo_try_verify_hosts, which caused the
4188 verification to occur when the EHLO/HELO command was issued. The ACL just
4189 tested the remembered result. Now, if a previous verification attempt has
4190 not happened, "verify = helo" does it there and then.
4192 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4194 TK/03 Fix log output including CR from clamd.
4196 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4197 exist provoked a memory error which could cause a segfault.
4199 PH/15 Installed PCRE 6.2
4201 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4203 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4204 of the problem. Specifically, suggested +O2 rather than +O1 for the
4207 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4209 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4210 record, the port was not correctly logged when the outgoing_port log
4211 selector was set (it logged the transort's default port).
4213 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4214 fallback_hosts, and "hosts" in the smtp transport.
4216 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4217 host errors such as "Connection refused".
4219 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4220 authentication with radiusclient 0.4.9:
4222 - Error returned from rc_read_config was caught wrongly
4223 - Username/password not passed on to radius server due to wrong length.
4225 The presumption is that some radiusclient API changes for 4.51/PH/17
4226 were not taken care of correctly. The code is still untested by me (my
4227 Linux distribution still has 0.3.2 of radiusclient), but it was
4228 contributed by a Radius user.
4230 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4231 expanding the "port" option of the smtp transport.
4233 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4234 while reading a MIME header. Thanks to Tom Hughes for a patch.
4236 PH/24 Include config.h inside local_scan.h so that configuration settings are
4239 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4240 that in an ACL for RCPT (for example), you can examine exactly what was
4243 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4244 commands, but it was not correctly comparing the address with the actual
4245 client host address. Thus, it would show the EHLO address in Received:
4246 header lines when this was not necessary.
4248 PH/27 Added the % operator to ${eval:}.
4250 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4251 it should be ignoring failures (because with -C, for example, it has lost
4252 privilege). It wasn't ignoring creation failures other than "already
4255 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4258 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4259 because an input line was too long, either on its own, or by virtue of
4260 too many continuations, the temporary file was not being removed, and the
4261 return code was incorrect.
4263 PH/31 Missing "BOOL" in function definition in filtertest.c.
4265 PH/32 Applied Sieve patches from the maintainer.
4267 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4269 PH/33 Added "verify = not_blind".
4271 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4272 Local/Makefile (with some defaults set). These are used in built scripts
4273 such as exicyclog, but they have never been used in the exim_install
4274 script (though there are many overriding facilities there). I have
4275 arranged that the exim_install script now takes note of these two
4278 PH/35 Installed configuration files for Dragonfly.
4280 PH/36 When a locally submitted message by a trusted user did not contain a
4281 From: header, and the sender address was obtained from -f or from an SMTP
4282 MAIL command, and the trusted user did not use -F to supply a sender
4283 name, $originator_name was incorrectly used when constructing a From:
4284 header. Furthermore, $originator_name was used for submission mode
4285 messages from external hosts without From: headers in a similar way,
4286 which is clearly wrong.
4288 PH/37 Added control=suppress_local_fixups.
4290 PH/38 When log_selector = +received_sender was set, and the addition of the
4291 sender made the log line's construction buffer exactly full, or one byte
4292 less than full, an overflow happened when the terminating "\n" was
4295 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4296 when the result of a list match is failure because a DNS lookup failed.
4298 PH/40 RM_COMMAND is now used in the building process.
4300 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4301 the "build-* directories that it finds.
4303 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4304 address in a domain literal was a prefix of an interface address.
4306 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4307 when verifying a sender address, unless rewrite_headers is false.
4309 PH/44 Wrote a long comment about why errors_to addresses are verified as
4310 recipients, not senders.
4312 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4313 the ratelimit ACL was added.
4315 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4317 PH/46 Added extra information about PostgreSQL errors to the error string.
4319 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4320 avoiding the need to install special zones in a real server. This is
4321 backwards compatible; if it can't find the fake resolver, it drops back.
4322 Thus, both old and new test suites can be run.
4324 TF/02 Added util/ratelimit.pl
4326 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4327 clock is set back in time.
4329 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4330 Candler <B.Candler@pobox.com>.
4332 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4333 OK for addresses that are the result of redirections.
4335 PH/48 A number of further additions for the benefit of the new test suite,
4336 including a fake gethostbyname() that interfaces to the fake DNS resolver
4339 TF/06 The fix for widen_domains has also been applied to qualify_single and
4340 search_parents which are the other dnslookup options that can cause
4343 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4346 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4348 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4349 service" if it encountered an MX record with an empty target hostname.
4350 The message is now "an MX or SRV record indicated no SMTP service".
4352 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4353 if the DNS of the sending site is misconfigured. This is quite a
4354 common situation. This change restores the behaviour of treating a
4355 helo verification defer as a failure.
4357 PH/51 If self=fail was set on a router, the bounce message did not include the
4358 actual error message.
4364 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4366 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4367 command itself ended in error, the underlying message about the transport
4368 filter timeout was being overwritten with the pipe command error. Now the
4369 underlying error message should be appended to the second error message.
4371 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4373 PH/02 Reduce the amount of output that "make" produces by default. Full output
4374 can still be requested.
4376 PH/03 The warning log line about a condition test deferring for a "warn" verb
4377 was being output only once per connection, rather than after each
4378 occurrence (because it was using the same function as for successful
4379 "warn" verbs). This seems wrong, so I have changed it.
4381 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4382 it should not have, which might have caused a crash in the right
4383 circumstances, but probably never did.
4385 PH/04 Installed a modified version of Tony Finch's patch to make submission
4386 mode fix the return path as well as the Sender: header line, and to
4387 add a /name= option so that you can make the user's friendly name appear
4390 TF/03 Added the control = fakedefer ACL modifier.
4392 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4393 Mark Lowes for thorough testing.
4395 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4397 TK/03 Merged latest SRS patch from Miles Wilton.
4399 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4400 with the definition in sysexits.h (which is #included earlier).
4401 Fortunately, Exim does not actually use EX_OK. The code used to try to
4402 preserve the sysexits.h value, by assuming that macro definitions were
4403 scanned for macro replacements. I have been disabused of this notion,
4404 so now the code just undefines EX_OK before #including unistd.h.
4406 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4407 in the smtp transport. When a block could not be written in a single
4408 write() function, the timeout was being re-applied to each part-write.
4409 This seems wrong - if the receiver was accepting one byte at a time it
4410 would take for ever. The timeout is now adjusted when this happens. It
4411 doesn't have to be particularly precise.
4413 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4414 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4416 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4417 without a domain if the check to <postmaster@domain> fails.
4419 SC/01 Eximstats: added -xls and the ability to specify output files
4420 (patch written by Frank Heydlauf).
4422 SC/02 Eximstats: use FileHandles for outputting results.
4424 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4426 SC/04 Eximstats: fixed display of large numbers with -nvr option
4428 SC/05 Eximstats: fixed merging of reports with empty tables.
4430 SC/06 Eximstats: added the -include_original_destination flag
4432 SC/07 Eximstats: removed tabs and trailing whitespace.
4434 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4436 TK/06 MBOX spool code: Add real "From " MBOX separator line
4437 so the .eml file is really in mbox format (even though
4438 most programs do not really care). Patch from Alex Miller.
4440 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4441 The latter is generated from $received_to and is only set if the
4442 message has one envelope recipient. SA can use these headers,
4443 obviously out-of-the-box. Patch from Alex Miller.
4445 PH/08 The ${def test on a variable was returning false if the variable's
4446 value was "0", contrary to what the specification has always said!
4447 The result should be true unless the variable is empty.
4449 PH/09 The syntax error of a character other than { following "${if
4450 def:variable_name" (after optional whitespace) was not being diagnosed.
4451 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4452 accidental colon was present, for example, could give incorrect results.
4454 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4455 result is used (not including appendfile, where other changes are about
4458 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4459 This involved changing a lot of size variables from int to off_t. It
4460 should work with maildirs and everything.
4462 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4463 spamd dying while we are connected to it.
4465 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4468 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4469 function for BDB 4.3.
4471 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4473 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4474 the sources for PCRE have been re-arranged and more files are now
4477 PH/15 The code I had for printing potentially long long variables in PH/11
4478 above was not the best (it lost precision). The length of off_t variables
4479 is now inspected at build time, and an appropriate printing format (%ld
4480 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4481 to be "long long int" or "long int". This is needed for the internal
4482 formatting function string_vformat().
4484 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4485 the configuration file to be ":syslog", then the script "guesses" where
4486 the logs files are, rather than using the compiled in default. In our
4487 case the guess is not the same as the compiled default, so the script
4488 suddenly stopped working when I started to use syslog. The patch checks
4489 to see if log_file_path is "". If so, it attempts to read it from exim
4490 with no configuration file to get the compiled in version, before it
4491 falls back to the previous guessing code."
4493 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4494 implementing BATV in an Exim configuration. See NewStuff for the gory
4497 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4498 Makefile that are specific to HP-UX.
4500 PH/18 If the "use_postmaster" option was set for a recipient callout together
4501 with the "random" option, the postmaster address was used as the MAIL
4502 FROM address for the random test, but not for the subsequent recipient
4503 test. It is now used for both.
4505 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4506 patch removes a few documentation additions to RFC 3028, because the
4507 latest draft now contains them. It adds the new en;ascii-case comparator
4508 and a new error check for 8bit text in MIME parts. Comparator and
4509 require names are now matched exactly. I enabled the subaddress
4510 extension, but it is not well tested yet (read: it works for me)."
4512 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4513 rework some of the code of TK/09 above to avoid the hardwired use of
4514 "%lld" and "long long". Replaced the call to snprintf() with a call to
4517 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4518 records point to non-existent hosts", "retry timeout exceeded", and
4519 "retry time not reached for any host after a long failure period".
4521 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4522 experimental DomainKeys support:
4524 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4525 (2) On an error such as an illegally used "control", the wrong name for
4526 the control was given.
4528 These problems did NOT occur unless DomainKeys support was compiled.
4530 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4532 PH/24 Added ${if match_ip condition.
4534 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4535 databases so that it will be absolutely obvious if a crash occurs in the
4536 DB library. This is a regular occurrence (often caused by mis-matched
4539 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4540 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4541 picked up on a user's system that detects such things. There doesn't seem
4542 to be a gcc warning option for this - only an attribute that has to be
4543 put on the function's prototype. It seems that in Fedora Core 4 they have
4544 set this on a number of new functions. No doubt there will be more in due
4547 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4548 specify a transport. However, if an address that was verified by such a
4549 router was the subject of a callout, Exim crashed because it tried to
4550 read the rcpt_include_affixes from the non-existent transport. Now it
4551 just assumes that the setting of that option is false. This bug was
4552 introduced by 4.51/PH/31.
4554 PH/28 Changed -d+all to exclude +memory, because that information is very
4555 rarely of interest, but it makes the output a lot bigger. People tend to
4556 do -d+all out of habit.
4558 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4559 code in os-type was giving problems when libc.so lives in lib64, like on
4562 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4563 aren't the modern standard, and it seems that some systems' include files
4564 don't always have them. Exim was already checking for some of the newer
4565 ones like T_AAAA, and defining it itself. I've added checks for all the
4566 record types that Exim uses.
4568 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4569 not automatically generating a new one, as it is supposed to. This
4570 prevented TLS from working. If the file did exist, but contained invalid
4571 data, a new version was generated, as expected. It was only the case of a
4572 non-existent file that was broken.
4574 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4575 with a change in libdomainkeys > 0.64.
4577 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4578 from DNS. If the selector record carries the flag, it now has
4579 precedence over the domain-wide flag.
4581 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4583 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4584 the use of a port name that isn't defined in /etc/services) occurred, the
4585 message was deferred as in a normal delivery, and thus remained on the
4586 spool, instead of being failed because of the mua_wrapper setting. This
4587 is now fixed, and I tidied up some of the mua_wrapper messages at the
4590 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4591 the messages in a hash of arrays rather than using individual hashes.
4592 This is a bit cleaner and results in dramatic memory savings, albeit
4593 at a slight CPU cost.
4595 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4596 as requested by Marc Sherman.
4598 SC/10 Eximstats: added histograms for user specified patterns as requested
4601 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4603 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4604 fopen() in the content-scanning modules that did not already have it.
4610 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4611 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4613 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4615 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4617 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4618 to test Sieve filters that use "vacation".
4620 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4621 that changes the way the GnuTLS parameters are stored in the cache file.
4622 The new format can be generated externally. For backward compatibility,
4623 if the data in the cache doesn't make sense, Exim assumes it has read an
4624 old-format file, and it generates new data and writes a new file. This
4625 means that you can't go back to an older release without removing the
4628 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4629 work if there are any delivery delays because "one_time" forces the
4630 parent to be marked "delivered", so its unseen clone is never tried
4631 again. For this reason, Exim now forbids the simultaneous setting of
4634 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4635 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4636 entry if you want to know the details. The fix, however, neglected to
4637 consider the case when local delivery batching is involved. The test for
4638 "previously delivered" was not happening when checking to see if an
4639 address could be batched with a previous (undelivered) one; under
4640 certain circumstances this could lead to multiple deliveries to the same
4643 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4644 in its include files, and this causes problems building Exim.
4646 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4647 header_syntax) but Exim was just ignoring anything given after a slash.
4648 In particular, this caused confusion with an attempt to use "verify =
4649 reverse_host_lookup/defer_ok". An error is now given when options are
4650 supplied for verify items that do not have them. (Maybe reverse_host_
4651 lookup should have a defer_ok option, but that's a different point.)
4653 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4654 defined by RFC 821) to 2048, because there were problems with some AUTH
4655 commands, and RFC 1869 says the size should be increased for extended
4656 SMTP commands that take arguments.
4658 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4661 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4662 "unknown" error; now it says that the functionality isn't in the binary.
4664 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4665 an address' error message when a string expansion fails (syntax or
4666 whatever). Otherwise the password may appear in the log. Following change
4667 PH/42 below, there is no longer a chance of it appearing in a bounce
4670 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4672 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4673 of its addresses was ever tried. (Bugzilla bug #2.)
4675 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4676 the result incorrectly in the debug output. (It correctly added a newline
4677 to what was transported.)
4679 TF/01 Added $received_time.
4681 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4682 commented out examples of how to interface to a virus scanner and to
4683 SpamAssassin. Also added commented examples of av_scanner and
4684 spamd_address settings.
4686 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4687 and controls are allowed in which ACLs. There were a couple of minor
4688 errors. Some of the entries in the conditions table (which is a table of
4689 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4690 negation of where the condition IS allowed.
4692 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4694 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4695 header file does not have a version number, so I've had to invent a new
4696 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4697 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4698 radiusclient), but it was contributed by a Radius user.
4700 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4701 files or directories, for OpenSSL.
4703 PH/19 When an Exim process that is running as root has to create an Exim log
4704 file, it does so in a subprocess that runs as exim:exim so as to get the
4705 ownership right at creation (otherwise, other Exim processes might see
4706 the file with the wrong ownership). There was no test for failure of this
4707 fork() call, which would lead to the process getting stuck as it waited
4708 for a non-existent subprocess. Forks do occasionally fail when resources
4709 run out. I reviewed all the other calls to fork(); they all seem to check
4712 PH/20 When checking for unexpected SMTP input at connect time (before writing
4713 the banner), Exim was not dealing correctly with a non-positive return
4714 from the read() function. If the client had disconnected by this time,
4715 the result was a log entry for a synchronization error with an empty
4716 string after "input=" when read() returned zero. If read() returned -1
4717 (an event I could not check), uninitialized data bytes were printed.
4718 There were reports of junk text (parts of files, etc) appearing after
4721 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4723 PH/22 Added support for macro redefinition, and (re)definition in between
4724 driver and ACL definitions.
4726 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4727 forgetting to use the resulting value; it was using the unexpanded value.
4729 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4730 hadn't been configured. The fix is from Juergen Kreileder, who
4731 understands it better than I do:
4733 "Here's what I see happening with three configured cyrus_sasl
4734 authenticators configured (plain, login, cram-md5):
4736 On startup auth_cyrus_sasl_init() gets called for each of these.
4737 This means three calls to sasl_listmech() without a specified mech_list.
4738 => SASL tests which mechs of all available mechs actually work
4739 => three warnings about OTP not working
4740 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4742 With the patch, sasl_listmech() also gets called three times. But now
4743 SASL's mech_list option is set to the server_mech specified in the the
4744 authenticator. Or in other words, the answer from sasl_listmech()
4745 gets limited to just the mech you're testing for (which is different
4747 => the return list contains just 'plain' or 'login', 'cram-md5' or
4748 nothing depending on the value of ob->server_mech.
4750 I've just tested the patch: Authentication still works fine,
4751 unavailable mechs specified in the exim configuration are still
4752 caught, and the auth.log warnings about OTP are gone."
4754 PH/25 When debugging is enabled, the contents of the command line are added
4755 to the debugging output, even when log_selector=+arguments is not
4758 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4759 answer is "GNU", and only if the return is "GNU/something" is the answer
4762 PH/27 $acl_verify_message is now set immediately after the failure of a
4763 verification in an ACL, and so is available in subsequent modifiers. In
4764 particular, the message can be preserved by coding like this:
4766 warn !verify = sender
4767 set acl_m0 = $acl_verify_message
4769 Previously, $acl_verify_message was set only while expanding "message"
4770 and "log_message" when a very denied access.
4772 PH/28 Modified OS/os.c-Linux with
4774 -#ifndef OS_LOAD_AVERAGE
4775 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4777 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4778 nomenclature these days.)
4780 PH/29 Installed patch from the Sieve maintainer that adds the options
4781 sieve_useraddress and sieve_subaddress to the redirect router.
4783 PH/30 In these circumstances:
4784 . Two addresses routed to the same list of hosts;
4785 . First host does not offer TLS;
4786 . First host accepts first address;
4787 . First host gives temporary error to second address;
4788 . Second host offers TLS and a TLS session is established;
4789 . Second host accepts second address.
4790 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4791 and peerdn, if requested) that were in fact used only for the second
4794 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4795 attention to any local part prefix or suffix that was matched by the
4796 router that accepted the address. It now behaves in the same way as it
4797 does for delivery: the affixes are removed from the local part unless
4798 rcpt_include_affixes is set on the transport.
4800 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4801 timeout during the DATA phase of an incoming message.
4803 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4804 applied a patch sanctioned by the Sieve maintainer.
4806 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4807 the uid or gid is negative. A case of a negative gid caused this to be
4808 noticed. The fix allows for either to be negative.
4810 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4811 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4814 PH/36 The change PH/12 above was broken. Fixed it.
4816 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4817 the grounds that routing the same address twice would always produce the
4818 same answer. This might have been true once, but it is certainly no
4819 longer true now. Routing a child address may depend on the previous
4820 routing that produced that child. Some complicated redirection strategies
4821 went wrong when messages had multiple recipients, and made Exim's
4822 behaviour dependent on the order in which the addresses were given.
4824 I have moved the duplicate checking until after the routing is complete.
4825 Exim scans the addresses that are assigned to local and remote
4826 transports, and removes any duplicates. This means that more work will be
4827 done, as duplicates will always all be routed, but duplicates are
4828 presumably rare, so I don't expect this is of any significance.
4830 For deliveries to pipes, files, and autoreplies, the duplicate checking
4831 still happens during the routing process, since they are not going to be
4834 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4835 It corrects a timeout issue with spamd. This is Ian's comment: "The
4836 background is that sometimes spamd either never reads data from a
4837 connection it has accepted, or it never writes response data. The exiscan
4838 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4839 blindly assumes that writes won't block so it may never time out."
4841 PH/39 Allow G after quota size as well as K and M.
4843 PH/40 The value set for $authenticated_id in an authenticator may not contain
4844 binary zeroes or newlines because the value is written to log lines and
4845 to spool files. There was no check on this. Now the value is run through
4846 the string_printing() function so that such characters are converted to
4847 printable escape sequences.
4849 PH/41 $message_linecount is a new variable that contains the total number of
4850 lines in the message. Compare $body_linecount, which is the count for the
4853 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4854 bounce and delay warning messages, except in certain special cases, which
4857 (a) An SMTP error message from a remote host;
4858 (b) A message specified in a :fail: redirection;
4859 (c) A message specified in a "fail" command in a system filter;
4860 (d) A message specified in a FAIL return from the queryprogram router;
4861 (e) A message specified by the cannot_route_message router option.
4863 In these cases only, Exim does include the error details in bounce and
4864 warning messages. There are also a few cases where bland messages such
4865 as "unrouteable address" or "local delivery error" are given.
4867 PH/43 $value is now also set for the "else" part of a ${run expansion.
4869 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4870 being worked on, but at least Exim now implements the latest version to
4873 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4874 process to complete was treated as a delivery failure, a timeout while
4875 writing the message to the pipe was logged, but erroneously treated as a
4876 successful delivery. Such timeouts include transport filter timeouts. For
4877 consistency with the overall process timeout, these timeouts are now
4878 treated as errors, giving rise to delivery failures by default. However,
4879 there is now a new Boolean option for the pipe transport called
4880 timeout_defer, which, if set TRUE, converts the failures into defers for
4881 both kinds of timeout. A transport filter timeout is now identified in
4884 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4885 systems where "make" and "gmake" are different, calling "gmake" at top
4886 level broke things. I've arranged for the value of $(MAKE) to be passed
4887 from the Makefile to this script so that it can call the same version of
4891 A note about Exim versions 4.44 and 4.50
4892 ----------------------------------------
4894 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4895 changes of various kinds. As a consequence, a big documentation update was
4896 needed. This delayed the release for rather longer than seemed good, especially
4897 in the light of a couple of (minor) security issues. Therefore, the changes
4898 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4899 release. So 4.44 and 4.50 are in effect two different branches that both start
4902 I have left the 4.50 change log unchanged; it contains all the changes since
4903 4.43. The change log for 4.44 is below; many of its items are identical to
4904 those for 4.50. This seems to be the most sensible way to preserve the
4905 historical information.
4911 1. Minor wording change to the doc/README.SIEVE file.
4913 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4914 computation of the current number of files was incorrect.
4916 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4917 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4920 4. Give more explanation in the error message when the command for a transport
4921 filter fails to execute.
4923 5. There are several places where Exim runs a non-Exim command in a
4924 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4925 was being done only for the command run by the queryprogram router. It is
4926 now done for all such subprocesses. The other cases are: ${run, transport
4927 filters, and the commands run by the lmtp and pipe transports.
4929 6. Added CONFIGURE_GROUP build-time option.
4931 7. Some older OS have a limit of 256 on the maximum number of file
4932 descriptors. Exim was using setrlimit() to set 1000 as a large value
4933 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4934 systems. I've change it so that if it can't get 1000, it tries for 256.
4936 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4937 was an oversight, and furthermore, ever since the addition of extra
4938 controls (e.g. 4.43/32), the checks on when to allow different forms of
4939 "control" were broken. There should now be diagnostics for all cases when a
4940 control that does not make sense is encountered.
4942 9. Added the /retain_sender option to "control=submission".
4944 10. $recipients is now available in the predata ACL (oversight).
4946 11. Tidy the search cache before the fork to do a delivery from a message
4947 received from the command line. Otherwise the child will trigger a lookup
4948 failure and thereby defer the delivery if it tries to use (for example) a
4949 cached ldap connection that the parent has called unbind on.
4951 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4952 of $address_data from the recipient verification was clobbered by the
4953 sender verification.
4955 13. The value of address_data from a sender verification is now available in
4956 $sender_address_data in subsequent conditions in the ACL statement.
4958 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4960 15. Added a new option "connect=<time>" to callout options, to set a different
4963 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4964 was its contents. (It was OK if the option was not defined at all.)
4966 17. A "Completed" log line is now written for messages that are removed from
4967 the spool by the -Mrm option.
4969 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4970 information about exactly what failed.
4972 19. Added -dd to debug only the daemon process.
4974 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4975 handles timeouts, both on the server side and network timeouts. Renamed the
4976 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4978 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4979 It is now set to "smtps".
4981 22. $host_address is now set to the target address during the checking of
4982 ignore_target_hosts.
4984 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4985 being passed; this would have caused $sender_host_name to have been used if
4986 matching the list had actually called for a host name (not very likely,
4987 since this list is usually IP addresses). A host name is now passed as
4990 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4991 code that specifies a non-restarting handler (typically sigaction() in
4992 modern systems) in an attempt to fix a rare and obscure crash bug.
4994 25. Narrowed the window for a race in the daemon that could cause it to ignore
4995 SIGCHLD signals. This is not a major problem, because they are used only to
4996 wake it up if nothing else does.
4998 26. A malformed maildirsize file could cause Exim to calculate negative values
4999 for the mailbox size or file count. Odd effects could occur as a result.
5000 The maildirsize information is now recalculated if the size or filecount
5003 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5004 support for a long time. Removed HAVE_SYS_VFS_H.
5006 28. Installed the latest version of exipick from John Jetmore.
5008 29. In an address list, if the pattern was not a regular expression, an empty
5009 subject address (from a bounce message) matched only if the pattern was an
5010 empty string. Non-empty patterns were not even tested. This was the wrong
5011 because it is perfectly reasonable to use an empty address as part of a
5012 database query. An empty address is now tested by patterns that are
5013 lookups. However, all the other forms of pattern expect the subject to
5014 contain a local part and a domain, and therefore, for them, an empty
5015 address still always fails if the pattern is not itself empty.
5017 30. Exim went into a mad DNS loop when attempting to do a callout where the
5018 host was specified on an smtp transport, and looking it up yielded more
5019 than one IP address.
5021 31. Re-factored the code for checking spool and log partition space into a
5022 function that finds that data and another that does the check. The former
5023 is then used to implement four new variables: $spool_space, $log_space,
5024 $spool_inodes, and $log_inodes.
5026 32. The RFC2047 encoding function was originally intended for short strings
5027 such as real names; it was not keeping to the 75-character limit for
5028 encoded words that the RFC imposes. It now respects the limit, and
5029 generates multiple encoded words if necessary. To be on the safe side, I
5030 have increased the buffer size for the ${rfc2047: expansion operator from
5033 33. It is now permitted to omit both strings after an "if" condition; if the
5034 condition is true, the result is "true". As before, when the second string
5035 is omitted, a false condition yields an empty string. This makes it less
5036 cumbersome to write custom ACL and router conditions.
5038 34. Failure to deliver a bounce message always caused it to be frozen, even if
5039 there was an errors_to setting on the router. The errors_to setting is now
5042 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5043 canonical form (fully expanded) before being placed in
5044 $sender_host_address.
5046 36. The table in the code that translates DNS record types into text (T_A to
5047 "A" for instance) was missing entries for NS and CNAME. It is just possible
5048 that this could have caused confusion if both these types were looked up
5049 for the same domain, because the text type is used as part of Exim's
5050 per-process caching. But the chance of anyone hitting this buglet seems
5053 37. The dnsdb lookup has been extended in a number of ways.
5055 (1) There is a new type, "zns", which walks up the domain tree until it
5056 finds some nameserver records. It should be used with care.
5058 (2) There is a new type, "mxh", which is like "mx" except that it returns
5059 just the host names, not the priorities.
5061 (3) It is now possible to give a list of domains (or IP addresses) to be
5062 looked up. The behaviour when one of the lookups defers can be
5063 controlled by a keyword.
5065 (4) It is now possible to specify the separator character for use when
5066 multiple records are returned.
5068 38. The dnslists ACL condition has been extended: it is now possible to supply
5069 a list of IP addresses and/or domains to be looked up in a particular DNS
5072 39. Added log_selector=+queue_time_overall.
5074 40. When running the queue in the test harness, wait just a tad after forking a
5075 delivery process, to get repeatability of debugging output.
5077 41. Include certificate and key file names in error message when GnuTLS fails
5078 to set them up, because the GnuTLS error message doesn't include the name
5079 of the failing file when there is a problem reading it.
5081 42. Allow both -bf and -bF in the same test run.
5083 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5085 44. The "Exiscan patch" is now merged into the mainline Exim source.
5087 45. Sometimes the final signoff response after QUIT could fail to get
5088 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5089 < 0 before doing a fflush(). This bug looks as though it goes back to the
5090 introduction of TLS in release 3.20, but "sometimes" must have been rare
5091 because the tests only now provoked it.
5093 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5094 (this can affect the format of dates).
5096 47. exim_tidydb, when checking for the continued existence of a message for
5097 which it has found a message-specific retry record, was not finding
5098 messages that were in split spool directories. Consequently, it was
5099 deleting retry records that should have stayed in existence.
5101 48. Steve fixed some bugs in eximstats.
5103 49. The SPA authentication driver was not abandoning authentication and moving
5104 on to the next authenticator when an expansion was forced to fail,
5105 contradicting the general specification for all authenticators. Instead it
5106 was generating a temporary error. It now behaves as specified.
5108 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5109 (the order specifies the preference for clients). The order is now AES256,
5110 AES128, 3DES, ARCFOUR128.
5112 51. Small patch to Sieve code - explicitly set From: when generating an
5115 52. Exim crashed if a remote delivery caused a very long error message to be
5116 recorded - for instance if somebody sent an entire SpamAssassin report back
5117 as a large number of 550 error lines. This bug was coincidentally fixed by
5118 increasing the size of one of Exim's internal buffers (big_buffer) that
5119 happened as part of the Exiscan merge. However, to be on the safe side, I
5120 have made the code more robust (and fixed the comments that describe what
5123 53. Now that there can be additional text after "Completed" in log lines (if
5124 the queue_time_overall log selector is set), a one-byte patch to exigrep
5125 was needed to allow it to recognize "Completed" as not the last thing in
5128 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5129 patch that reportedly fixes this has been added. I am not expert enough to
5130 create a test for it. This is what the patch creator wrote:
5132 "I found a little strange behaviour of ldap code when working with
5133 Windows 2003 AD Domain, where users was placed in more than one
5134 Organization Units. When I tried to give exim partial DN, the exit code
5135 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5136 But simultaneously result of request was absolutely normal ldap result,
5137 so I produce this patch..."
5139 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5140 REFERENCE, so I have modified the code to exclude the patch when that macro
5143 55. Some experimental protocols are using DNS PTR records for new purposes. The
5144 keys for these records are domain names, not reversed IP addresses. The
5145 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5146 leaves it alone. Component reversal etc. now happens only for IP addresses.
5149 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5151 57. Double the size of the debug message buffer (to 2048) so that more of very
5152 long debug lines gets shown.
5154 58. The exicyclog utility now does better if the number of log files to keep
5155 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5157 59. Two changes related to the smtp_active_hostname option:
5159 (1) $smtp_active_hostname is now available as a variable.
5160 (2) The default for smtp_banner uses $smtp_active_hostname instead
5161 of $primary_hostname.
5163 60. The host_aton() function is supposed to be passed a string that is known
5164 to be a valid IP address. However, in the case of IPv6 addresses, it was
5165 not checking this. This is a hostage to fortune. Exim now panics and dies
5166 if the condition is not met. A case was found where this could be provoked
5167 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5168 components; fortuitously, this particular loophole had already been fixed
5169 by change 4.50/55 above.
5171 If there are any other similar loopholes, the new check in host_aton()
5172 itself should stop them being exploited. The report I received stated that
5173 data on the command line could provoke the exploit when Exim was running as
5174 exim, but did not say which command line option was involved. All I could
5175 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5176 running as the user.
5179 61. There was a buffer overflow vulnerability in the SPA authentication code
5180 (which came originally from the Samba project). I have added a test to the
5181 spa_base64_to_bits() function which I hope fixes it.
5184 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5185 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5187 63. The daemon start-up calls getloadavg() while still root for those OS that
5188 need the first call to be done as root, but it missed one case: when
5189 deliver_queue_load_max is set with deliver_drop_privilege. This is
5190 necessary for the benefit of the queue runner, because there is no re-exec
5191 when deliver_drop_privilege is set.
5193 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5194 This has been fixed.
5196 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5197 was in use, was not putting the data itself into the right store pool;
5198 consequently, it could be overwritten for a subsequent message in the same
5199 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5202 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5204 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5205 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5206 the function were treating the return as a boolean value, which happened to
5207 work because 0=false and not-0=true, but is not correct code.
5209 68. The host_aton() function was not handling scoped IPv6 addresses (those
5210 with, for example, "%eth0" on the end) correctly.
5212 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5213 negated items (that is, ~something) in unsigned ints. Some compilers
5214 apparently mutter when there is no cast.
5216 70. If an address verification called from an ACL failed, and did not produce a
5217 user-specific message (i.e. there was only a "system" message), nothing was
5218 put in $acl_verify_message. In this situation, it now puts the system
5221 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5222 session; change 4.31/43 added the unwanted input to the log line - except
5223 that it did not do this in the start of session case. It now does.
5225 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5226 This is wrong and can cause the other end to generate a synchronization
5227 error if it is another Exim or anything else that does the synchronization
5228 check. A QUIT command is no longer sent after a timeout.
5230 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5231 during host lookups.
5233 74. The defer_ok option of callout verification was not working if it was used
5234 when verifying addresses in header lines, that is, for this case:
5236 verify = header_sender/callout=defer_ok
5238 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5239 those file descriptors could be used for SMTP connections. If anything
5240 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5241 could be sent to the SMTP client, causing chaos. The daemon now opens
5242 stdin, stdout, and stderr to /dev/null when it puts itself into the
5245 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5246 log by default. The user can override this with suitable Perl magic.
5248 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5249 the log message when discard triggers, was not working for the DATA ACL or
5250 for the non-SMTP ACL.
5252 78. Error message wording change in sieve.c.
5254 79. If smtp_accept_max_per_host was set, the number of connections could be
5255 restricted to fewer than expected, because the daemon was trying to set up
5256 a new connection before checking whether the processes handling previous
5257 connections had finished. The check for completed processes is now done
5258 earlier. On busy systems, this bug wouldn't be noticed because something
5259 else would have woken the daemon, and it would have reaped the completed
5262 80. If a message was submitted locally by a user whose login name contained one
5263 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5264 It caused a spool format error. I have fixed the spool reading code. A
5265 related problem was that the "from" clause in the Received: line became
5266 illegal because of the space(s). It is now covered by ${quote_local_part.
5268 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5271 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5273 83. Patch from Sieve maintainer to fix unterminated string problem in
5274 "vacation" handling.
5276 84. Some minor changes to the Linux configuration files to help with other
5277 OS variants using glibc.
5279 85. One more patch for Sieve to update vacation handling to latest spec.
5282 ----------------------------------------------------
5283 See the note above about the 4.44 and 4.50 releases.
5284 ----------------------------------------------------
5290 1. Change 4.43/35 introduced a bug that caused file counts to be
5291 incorrectly computed when quota_filecount was set in an appendfile
5294 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5295 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5298 3. Give more explanation in the error message when the command for a transport
5299 filter fails to execute.
5301 4. There are several places where Exim runs a non-Exim command in a
5302 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5303 was being done only for the command run by the queryprogram router. It is
5304 now done for all such subprocesses. The other cases are: ${run, transport
5305 filters, and the commands run by the lmtp and pipe transports.
5307 5. Some older OS have a limit of 256 on the maximum number of file
5308 descriptors. Exim was using setrlimit() to set 1000 as a large value
5309 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5310 systems. I've change it so that if it can't get 1000, it tries for 256.
5312 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5313 was an oversight, and furthermore, ever since the addition of extra
5314 controls (e.g. 4.43/32), the checks on when to allow different forms of
5315 "control" were broken. There should now be diagnostics for all cases when a
5316 control that does not make sense is encountered.
5318 7. $recipients is now available in the predata ACL (oversight).
5320 8. Tidy the search cache before the fork to do a delivery from a message
5321 received from the command line. Otherwise the child will trigger a lookup
5322 failure and thereby defer the delivery if it tries to use (for example) a
5323 cached ldap connection that the parent has called unbind on.
5325 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5326 of $address_data from the recipient verification was clobbered by the
5327 sender verification.
5329 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5330 was its contents. (It was OK if the option was not defined at all.)
5332 11. A "Completed" log line is now written for messages that are removed from
5333 the spool by the -Mrm option.
5335 12. $host_address is now set to the target address during the checking of
5336 ignore_target_hosts.
5338 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5339 being passed; this would have caused $sender_host_name to have been used if
5340 matching the list had actually called for a host name (not very likely,
5341 since this list is usually IP addresses). A host name is now passed as
5344 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5345 code that specifies a non-restarting handler (typically sigaction() in
5346 modern systems) in an attempt to fix a rare and obscure crash bug.
5348 15. Narrowed the window for a race in the daemon that could cause it to ignore
5349 SIGCHLD signals. This is not a major problem, because they are used only to
5350 wake it up if nothing else does.
5352 16. A malformed maildirsize file could cause Exim to calculate negative values
5353 for the mailbox size or file count. Odd effects could occur as a result.
5354 The maildirsize information is now recalculated if the size or filecount
5357 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5358 support for a long time. Removed HAVE_SYS_VFS_H.
5360 18. Updated exipick to current release from John Jetmore.
5362 19. Allow an empty sender to be matched against a lookup in an address list.
5363 Previously the only cases considered were a regular expression, or an
5366 20. Exim went into a mad DNS lookup loop when doing a callout where the
5367 host was specified on the transport, if the DNS lookup yielded more than
5370 21. The RFC2047 encoding function was originally intended for short strings
5371 such as real names; it was not keeping to the 75-character limit for
5372 encoded words that the RFC imposes. It now respects the limit, and
5373 generates multiple encoded words if necessary. To be on the safe side, I
5374 have increased the buffer size for the ${rfc2047: expansion operator from
5377 22. Failure to deliver a bounce message always caused it to be frozen, even if
5378 there was an errors_to setting on the router. The errors_to setting is now
5381 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5382 canonical form (fully expanded) before being placed in
5383 $sender_host_address.
5385 24. Updated eximstats to version 1.33
5387 25. Include certificate and key file names in error message when GnuTLS fails
5388 to set them up, because the GnuTLS error message doesn't include the name
5389 of the failing file when there is a problem reading it.
5391 26. Expand error message when OpenSSL has problems setting up cert/key files.
5394 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5395 (this can affect the format of dates).
5397 28. exim_tidydb, when checking for the continued existence of a message for
5398 which it has found a message-specific retry record, was not finding
5399 messages that were in split spool directories. Consequently, it was
5400 deleting retry records that should have stayed in existence.
5402 29. eximstats updated to version 1.35
5403 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5404 1.35 - bugfix such that pie charts by volume are generated correctly
5406 30. The SPA authentication driver was not abandoning authentication and moving
5407 on to the next authenticator when an expansion was forced to fail,
5408 contradicting the general specification for all authenticators. Instead it
5409 was generating a temporary error. It now behaves as specified.
5411 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5412 (the order specifies the preference for clients). The order is now AES256,
5413 AES128, 3DES, ARCFOUR128.
5415 31. Small patch to Sieve code - explicitly set From: when generating an
5418 32. Exim crashed if a remote delivery caused a very long error message to be
5419 recorded - for instance if somebody sent an entire SpamAssassin report back
5420 as a large number of 550 error lines. This bug was coincidentally fixed by
5421 increasing the size of one of Exim's internal buffers (big_buffer) that
5422 happened as part of the Exiscan merge. However, to be on the safe side, I
5423 have made the code more robust (and fixed the comments that describe what
5426 33. Some experimental protocols are using DNS PTR records for new purposes. The
5427 keys for these records are domain names, not reversed IP addresses. The
5428 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5429 leaves it alone. Component reversal etc. now happens only for IP addresses.
5432 34. The host_aton() function is supposed to be passed a string that is known
5433 to be a valid IP address. However, in the case of IPv6 addresses, it was
5434 not checking this. This is a hostage to fortune. Exim now panics and dies
5435 if the condition is not met. A case was found where this could be provoked
5436 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5437 components; fortuitously, this particular loophole had already been fixed
5438 by change 4.50/55 or 4.44/33 above.
5440 If there are any other similar loopholes, the new check in host_aton()
5441 itself should stop them being exploited. The report I received stated that
5442 data on the command line could provoke the exploit when Exim was running as
5443 exim, but did not say which command line option was involved. All I could
5444 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5445 running as the user.
5448 35. There was a buffer overflow vulnerability in the SPA authentication code
5449 (which came originally from the Samba project). I have added a test to the
5450 spa_base64_to_bits() function which I hope fixes it.
5453 36. The daemon start-up calls getloadavg() while still root for those OS that
5454 need the first call to be done as root, but it missed one case: when
5455 deliver_queue_load_max is set with deliver_drop_privilege. This is
5456 necessary for the benefit of the queue runner, because there is no re-exec
5457 when deliver_drop_privilege is set.
5459 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5460 was in use, was not putting the data itself into the right store pool;
5461 consequently, it could be overwritten for a subsequent message in the same
5462 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5465 38. Sometimes the final signoff response after QUIT could fail to get
5466 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5467 < 0 before doing a fflush(). This bug looks as though it goes back to the
5468 introduction of TLS in release 3.20, but "sometimes" must have been rare
5469 because the tests only now provoked it.
5475 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5476 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5477 Now it returns an integer. A number of calls were still expecting a T/F
5478 return. Fortuitously, in all cases, the tests worked in OK situations,
5479 which is the norm. However, things would have gone wrong on any write
5480 failures on the smtp file descriptor. This function is used when sending
5481 messages over SMTP and also when doing verify callouts.
5483 2. When Exim is called to do synchronous delivery of a locally submitted
5484 message (the -odf or -odi options), it no longer closes stderr before doing
5487 3. Implemented the mua_wrapper option.
5489 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5491 5. Implemented the functions header_remove(), header_testname(),
5492 header_add_at_position(), and receive_remove_recipient(), and exported them
5495 6. If an ACL "warn" statement specified the addition of headers, Exim already
5496 inserted X-ACL-Warn: at the start if there was no header name. However, it
5497 was not making this test for the second and subsequent header lines if
5498 there were newlines in the string. This meant that an invalid header could
5499 be inserted if Exim was badly configured.
5501 7. Allow an ACL "warn" statement to add header lines at the start or after all
5502 the Received: headers, as well as at the end.
5504 8. Added the rcpt_4xx retry error code.
5506 9. Added postmaster_mailfrom=xxx to callout verification option.
5508 10. Added mailfrom=xxxx to the callout verification option, for verify=
5511 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5512 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5514 12. Inserted some casts to stop certain compilers warning when using pointer
5515 differences as field lengths or precisions in printf-type calls (mostly
5516 affecting debugging statements).
5518 13. Added optional readline() support for -be (dynamically loaded).
5520 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5521 same clock tick as a message's arrival, so that its received time was the
5522 same as the "first fail" time on the retry record, and that message
5523 remained on the queue past the ultimate address timeout, every queue runner
5524 would try a delivery (because it was past the ultimate address timeout) but
5525 after another failure, the ultimate address timeout, which should have then
5526 bounced the address, did not kick in. This was a "< instead of <=" error;
5527 in most cases the first failure would have been in the next clock tick
5528 after the received time, and all would be well.
5530 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5531 being recognized when the domain list was tested by the match_domain
5532 condition in an expansion string.
5534 16. Added the ${str2b64: operator.
5536 17. Exim was always calling setrlimit() to set a large limit for the number of
5537 processes, without checking whether the existing limit was already
5538 adequate. (It did check for the limit on file descriptors.) Furthermore,
5539 errors from getrlimit() and setrlimit() were being ignored. Now they are
5540 logged to the main and panic logs, but Exim does carry on, to try to do its
5541 job under whatever limits there are.
5543 18. Imported PCRE 5.0.
5545 19. Trivial typo in log message " temporarily refused connection" (the leading
5548 20. If the log selector return_path_on_delivery was set and an address was
5549 redirected to /dev/null, the delivery process crashed because it assumed
5550 that a return path would always be set for a "successful" delivery. In this
5551 case, the whole delivery is bypassed as an optimization, and therefore no
5554 21. Internal re-arrangement: the function for sending a challenge and reading
5555 a response while authentication was assuming a zero-terminated challenge
5556 string. It's now changed to take a pointer and a length, to allow for
5557 binary data in such strings.
5559 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5561 23. Exim was not respecting finduser_retries when seeking the login of the
5562 uid under which it was called; it was always trying 10 times. (The default
5563 setting of finduser_retries is zero.) Also, it was sleeping after the final
5564 failure, which is pointless.
5566 24. Implemented tls_on_connect_ports.
5568 25. Implemented acl_smtp_predata.
5570 26. If the domain in control=submission is set empty, Exim assumes that the
5571 authenticated id is a complete email address when it generates From: or
5572 Sender: header lines.
5574 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5575 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5576 chown and chgrp in /bin and hostname in /usr/bin.
5578 28. Exim was keeping the "process log" file open after each use, just as it
5579 does for the main log. This opens the possibility of it remaining open for
5580 long periods when the USR1 signal hits a daemon. Occasional processlog
5581 errors were reported, that could have been caused by this. Anyway, it seems
5582 much more sensible not to leave this file open at all, so that is what now
5585 29. The long-running daemon process does not normally write to the log once it
5586 has entered its main loop, and it closes the log before doing so. This is
5587 so that log files can straightforwardly be renamed and moved. However,
5588 there are a couple of unusual error situations where the daemon does write
5589 log entries, and I had neglected to close the log afterwards.
5591 30. The text of an SMTP error response that was received during a remote
5592 delivery was being truncated at 512 bytes. This is too short for some of
5593 the long messages that one sometimes sees. I've increased the limit to
5596 31. It is now possible to make retry rules that apply only when a message has a
5597 specific sender, in particular, an empty sender.
5599 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5600 it possible to be selective about when SMTP synchronization is enforced.
5602 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5604 32. Implemented hosts_connection_nolog.
5606 33. Added an ACL for QUIT.
5608 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5611 35. Added mailbox_size and mailbox_filecount to appendfile.
5613 36. Added control = no_multiline_responses to ACLs.
5615 37. There was a bug in the logic of the code that waits for the clock to tick
5616 in the case where the clock went backwards by a substantial amount such
5617 that the microsecond fraction of "now" was more than the microsecond
5618 fraction of "then" (but the whole seconds number was less).
5620 38. Added support for the libradius Radius client library this is found on
5621 FreeBSD (previously only the radiusclient library was supported).
5627 1. When certain lookups returned multiple values in the form name=value, the
5628 quoting of the values was not always being done properly. Specifically:
5629 (a) If the value started with a double quote, but contained no whitespace,
5631 (b) If the value contained whitespace other than a space character (i.e.
5632 tabs or newlines or carriage returns) it was not quoted.
5633 This fix has been applied to the mysql and pgsql lookups by writing a
5634 separate quoting function and calling it from the lookup code. The fix
5635 should probably also be applied to nisplus, ibase and oracle lookups, but
5636 since I cannot test any of those, I have not disturbed their existing code.
5638 2. A hit in the callout cache for a specific address caused a log line with no
5639 reason for rejecting RCPT. Now it says "Previous (cached) callout
5640 verification failure".
5642 3. There was an off-by-one bug in the queryprogram router. An over-long
5643 return line was truncated at 256 instead of 255 characters, thereby
5644 overflowing its buffer with the terminating zero. As well as fixing this, I
5645 have increased the buffer size to 1024 (and made a note to document this).
5647 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5648 when Exim is waiting for an SMTP response from a remote server, Exim
5649 restarts its select() call on the socket, thereby resetting its timeout.
5650 This is not a problem when such interrupts are rare. Somebody set up a cron
5651 job to run exiwhat every 2 minutes, which is less than the normal select()
5652 timeout (5 or 10 minutes). This meant that the select() timeout never
5653 kicked in because it was always reset. I have fixed this by comparing the
5654 time when an interrupt arrives with the time at the start of the first call
5655 to select(). If more time than the timeout has elapsed, the interrupt is
5656 treated as a timeout.
5658 5. Some internal re-factoring in preparation for the addition of Sieve
5659 extensions (by MH). In particular, the "personal" test is moved to a
5660 separate function, and given an option for scanning Cc: and Bcc: (which is
5661 not set for Exim filters).
5663 6. When Exim created an email address using the login of the caller as the
5664 local part (e.g. when creating a From: or Sender: header line), it was not
5665 quoting the local part when it contained special characters such as @.
5667 7. Installed new OpenBSD configuration files.
5669 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5670 try to make them clearer.
5672 9. Callout options, other than the timeout value, were being ignored when
5673 verifying sender addresses in header lines. For example, when using
5675 verify = header_sender/callout=no_cache
5677 the cache was (incorrectly) being used.
5679 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5680 only the Cygwin environment.
5682 11. When return_path_on_delivery was set as a log selector, if different remote
5683 addresses in the same message used different return paths and parallel
5684 remote delivery occurred, the wrong values would sometimes be logged.
5685 (Whenever a remote delivery process finished, the return path value from
5686 the most recently started remote delivery process was logged.)
5688 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5689 lines when AUTH and/or TLS are in use. This is the "received protocol"
5690 field. Exim used to use "asmtp" for authenticated SMTP, without any
5691 indication (in the protocol name) for TLS use. Now it follows the RFC and
5692 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5693 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5694 names appear in log lines as well as in Received: header lines.
5696 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5697 extensions, and comparison tests, and to fix some bugs.
5699 14. Changes to the "personal" filter test:
5701 (1) The test was buggy in that it was just doing the equivalent of
5702 "contains" tests on header lines. For example, if a user's address was
5703 anne@some.where, the "personal" test would incorrectly be true for
5705 To: susanne@some.where
5707 This test is now done by extracting each address from the header in turn,
5708 and checking the entire address. Other tests that are part of "personal"
5709 are now done using regular expressions (for example, to check local parts
5710 of addresses in From: header lines).
5712 (2) The list of non-personal local parts in From: addresses has been
5713 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5714 taken from the Sieve specification recommendations.
5716 (3) If the message contains any header line starting with "List-" it is
5717 treated as non-personal.
5719 (4) The test for "circular" in the Subject: header line has been removed
5720 because it now seems ill-conceived.
5722 15. Minor typos in src/EDITME comments corrected.
5724 16. Installed latest exipick from John Jetmore.
5726 17. If headers_add on a router specified a text string that was too long for
5727 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5728 of string_sprintf() is now avoided.
5730 18. $message_body_size was not set (it was always zero) when running the DATA
5731 ACL and the local_scan() function.
5733 19. For the "mail" command in an Exim filter, no default was being set for
5734 the once_repeat time, causing a random time value to be used if "once" was
5735 specified. (If the value happened to be <= 0, no repeat happened.) The
5736 default is now 0s, meaning "never repeat". The "vacation" command was OK
5737 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5738 (I found it when inspecting the code).
5740 20. There is now an overall timeout for performing a callout verification. It
5741 defaults to 4 times the callout timeout, which applies to individual SMTP
5742 commands during the callout. The overall timeout applies when there is more
5743 than one host that can be tried. The timeout is checked before trying the
5744 next host. This prevents very long delays if there are a large number of
5745 hosts and all are timing out (e.g. when the network connections are timing
5746 out). The value of the overall timeout can be changed by specifying an
5747 additional sub-option for "callout", called "maxwait". For example:
5749 verify = sender/callout=5s,maxwait=20s
5751 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5752 to the end before writing, but this should make it even safer).
5754 22. Exim was forgetting that it had advertised PIPELINING for the second and
5755 subsequent messages on an SMTP connection. It was also not resetting its
5756 memory on STARTTLS and an internal HELO.
5758 23. When Exim logs an SMTP synchronization error within a session, it now
5759 records whether PIPELINING has been advertised or not.
5761 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5762 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5763 rather than long int.
5765 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5767 26. Added the never_mail option to autoreply.
5773 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5774 crash if the getsockname() call failed; this can happen if a connection is
5775 closed very soon after it is established. The problem was simply in the
5776 order in which certain operations were done, causing Exim to try to write
5777 to the SMTP stream before it had set up the file descriptor. The bug has
5778 been fixed by making things happen in the correct order.
5784 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5785 before the connection was closed, thus losing the rejection response.
5787 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5788 some early Solaris releases, but causes trouble in current releases where
5789 socklen_t is defined.
5791 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5794 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5797 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5798 that releasing the top of it at the end releases what was used for sub-
5799 expansions (unless the block got too big). However, discard this block if
5800 the first thing is a variable or header, so that we can use its block when
5801 it is dynamic (useful for very large $message_headers, for example).
5803 6. Lookups now cache *every* query, not just the most recent. A new, separate
5804 store pool is used for this. It can be recovered when all lookup caches are
5805 flushed. Lookups now release memory at the end of their result strings.
5806 This has involved some general refactoring of the lookup sources.
5808 7. Some code has been added to the store_xxx() functions to reduce the amount
5809 of flapping under certain conditions.
5811 8. log_incoming_interface used to affect only the <= reception log lines. Now
5812 it causes the local interface and port to be added to several more SMTP log
5813 lines, for example "SMTP connection from", and rejection lines.
5815 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5817 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5819 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5820 could be overwritten at the end of the current message (or the start of a
5821 new message if it was set in a HELO ACL). The value is now preserved for
5822 the duration of the SMTP connection.
5824 12. If a transport had a headers_rewrite setting, and a matching header line
5825 contained an unqualified address, that address was qualified, even if it
5826 did not match any rewriting rules. The underlying bug was that the values
5827 of the flags that permit the existence of unqualified sender and recipient
5828 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5829 non-local messages, and by -bnq for local messages) were not being
5830 preserved with the message after it was received.
5832 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5833 "next input=" as part of the text comprising the host identity instead of
5834 the correct text. The code was using the same buffer for two different
5835 strings. However, depending on which order the printing function evaluated
5836 its arguments, the bug did not always show up. Under Linux, for example, my
5837 test suite worked just fine.
5839 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5840 doesn't work with some older versions of Perl. It has been changed to "my",
5841 which in any case is probably the better facility to use.
5843 15. A really picky compiler found some instances of statements for creating
5844 error messages that either had too many or two few arguments for the format
5847 16. The size of the buffer for calls to the DNS resolver has been increased
5848 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5849 for addresses that have a lot of PTR records. This alleviates a problem; it
5850 does not fully solve it.
5852 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5853 buffer now truncates the list and logs the incident, which is the same
5854 action as happens when Exim is looking up a host name and its aliases.
5855 Previously in this situation something unpredictable would happen;
5856 sometimes it was "internal error: store_reset failed".
5858 18. If a server dropped the connection unexpectedly when an Exim client was
5859 using GnuTLS and trying to read a response, the client delivery process
5860 crashed while trying to generate an error log message.
5862 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5863 string, for example:
5865 warn message = H1: something\nH2: something
5867 the text was added as a single header line from Exim's point of view
5868 though it ended up OK in the delivered message. However, searching for the
5869 second and subsequent header lines using $h_h2: did not work. This has been
5870 fixed. Similarly, if a system filter added multiple headers in this way,
5871 the routers could not see them.
5873 20. Expanded the error message when iplsearch is called with an invalid key to
5874 suggest using net-iplsearch in a host list.
5876 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5877 ACLs are no longer actually imposed (and a message to that effect is
5880 22. If a "gecos" field in a passwd entry contained escaped characters, in
5881 particular, if it contained a \" sequence, Exim got it wrong when building
5882 a From: or a Sender: header from that name. A second bug also caused
5883 incorrect handling when an unquoted " was present following a character
5884 that needed quoting.
5886 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5887 was not being matched caselessly.
5889 24. Arranged for all hyphens in the exim.8 source to be escaped with
5892 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5893 properly. Recipient callout cache records were still being keyed to include
5894 the sender, even when use_sender was set false. This led to far more
5895 callouts that were necessary. The sender is no longer included in the key
5896 when use_sender is false.
5898 26. Added "control = submission" modifier to ACLs.
5900 27. Added the ${base62d: operator to decode base 62 numbers.
5902 28. dnsdb lookups can now access SRV records.
5904 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5905 the configuration file.
5907 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5908 (-v) mode. This makes the output for a verbose queue run more intelligible.
5910 31. Added a use_postmaster feature to recipient callouts.
5912 32. Added the $body_zerocount variable, containing the number of binary zero
5913 bytes in the message body.
5915 33. The time of last modification of the "new" subdirectory is now used as the
5916 "mailbox time last read" when there is a quota error for a maildir
5919 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5921 35. Added +ignore_unknown as a special item in host lists.
5923 36. Code for decoding IPv6 addresses in host lists is now included, even if
5924 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5925 address was recognized as an IP address, but was then not correctly decoded
5926 into binary, causing unexpected and incorrect effects when compared with
5933 1. Very minor rewording of debugging text in manualroute to say "list of
5934 hosts" instead of "hostlist".
5936 2. If verify=header_syntax was set, and a header line with an unqualified
5937 address (no domain) and a large number of spaces between the end of the
5938 name and the colon was received, the reception process suffered a buffer
5939 overflow, and (when I tested it) crashed. This was caused by some obsolete
5940 code that should have been removed. The fix is to remove it!
5942 3. When running in the test harness, delay a bit after writing a bounce
5943 message to get a bit more predictability in the log output.
5945 4. Added a call to search_tidyup() just before forking a reception process. In
5946 theory, someone could use a lookup in the expansion of smtp_accept_max_
5947 per_host which, without the tidyup, could leave open a database connection.
5949 5. Added the variables $recipient_data and $sender_data which get set from a
5950 lookup success in an ACL "recipients" or "senders" condition, or a router
5951 "senders" option, similar to $domain_data and $local_part_data.
5953 6. Moved the writing of debug_print from before to after the "senders" test
5956 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5957 problems for message scanning, either using a data ACL, or using
5958 local_scan() because the Received: header was not generated till after they
5959 were called (in order to set the time as the time of reception completion).
5960 I have revised the way this works. The header is now generated after the
5961 body is received, but before the ACL or local_scan() are called. After they
5962 are run, the timestamp in the header is updated.
5968 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5969 before starting a queue runner without re-exec. This happened only when
5970 deliver_drop_privilege was set or when the Exim user was set to root. The
5971 effect of the bug was that timeouts during subsequent deliveries caused
5972 crashes instead of being properly handled. The handler is now left at its
5973 default (and expected) setting.
5975 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5976 message, again when deliver_drop_privilege is set or Exim is run as root.
5977 The bug described in (1) was not present in this case, but the tidying up
5978 of the other signals was missing. I have made the two cases consistent.
5980 3. The ignore_target_hosts setting on a manualroute router was being ignored
5981 for hosts that were looked up using the /MX notation.
5983 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5986 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5987 operated on the sender address. After changing the $sender_address to <>
5988 for the sender address verify, Exim was re-instated it as the original
5989 (before rewriting) address, but remembering that it had rewritten it, so it
5990 wasn't rewriting it again. This bug also had the effect of breaking the
5991 sender address verification caching when the sender address was rewritten.
5993 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5994 This has been changed so that if the ip literal address matches
5995 ignore_target_hosts, the router declines.
5997 7. Added expansion conditions match_domain, match_address, and match_local_
5998 part (NOT match_host).
6000 8. The placeholder for the Received: header didn't have a length field set.
6002 9. Added code to Exim itself and to exim_lock to test for a specific race
6003 condition that could lead to file corruption when using MBX delivery. The
6004 issue is with the lockfile that is created in /tmp. If this file is removed
6005 after a process has opened it but before that process has acquired a lock,
6006 there is the potential for a second process to recreate the file and also
6007 acquire a lock. This could lead to two Exim processes writing to the file
6008 at the same time. The added code performs the same test as UW imapd; it
6009 checks after acquiring the lock that its file descriptor still refers to
6010 the same named file.
6012 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6013 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6016 11. Added the smtp_active_hostname option. If used, this will typically be made
6017 to depend on the incoming interface address. Because $interface_address is
6018 not set up until the daemon has forked a reception process, error responses
6019 that can happen earlier (such as "too many connections") no longer contain
6022 12. If an expansion in a condition on a "warn" statement fails because a lookup
6023 defers, the "warn" statement is abandoned, and the next ACL statement is
6024 processed. Previously this caused the whole ACL to be aborted.
6026 13. Added the iplsearch lookup type.
6028 14. Added ident_timeout as a log selector.
6030 15. Added tls_certificate_verified as a log selector.
6032 16. Added a global option tls_require_ciphers (compare the smtp transport
6033 option of the same name). This controls incoming TLS connections.
6035 17. I finally figured out how to make tls_require_ciphers do a similar thing
6036 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6037 before starting the TLS session.
6039 18. Tabs are now shown as \t in -bP output.
6041 19. If the log selector return_path_on_delivery was set, Exim crashed when
6042 bouncing a message because it had too many Received: header lines.
6044 20. If two routers both had headers_remove settings, and the first one included
6045 a superfluous trailing colon, the final name in the first list and the
6046 first name in the second list were incorrectly joined into one item (with a
6047 colon in the middle).
6053 1. Added -C and -D options to the exinext utility, mainly to make it easier
6054 to include in the automated testing, but these could be helpful when
6055 multiple configurations are in use.
6057 2. The exinext utility was not formatting the output nicely when there was
6058 an alternate port involved in the retry record key, nor when there was a
6059 message id as well (for retries that were specific to a specific message
6060 and a specific host). It was also confused by IPv6 addresses, because of
6061 the additional colons they contain. I have fixed the IPv4 problem, and
6062 patched it up to do a reasonable job for IPv6.
6064 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6065 delivery, the log line now contains "pipelined" if PIPELINING was used.
6067 4. An SMTP transport process used to panic and die if the bind() call to set
6068 an explicit outgoing interface failed. This has been changed; it is now
6069 treated in the same way as a connect() failure.
6071 5. A reference to $sender_host_name in the part of a conditional expansion
6072 that was being skipped was still causing a DNS lookup. This no longer
6075 6. The def: expansion condition was not recognizing references to header lines
6076 that used bh_ and bheader_.
6078 7. Added the _cache feature to named lists.
6080 8. The code for checking quota_filecount in the appendfile transport was
6081 allowing one more file than it should have been.
6083 9. For compatibility with Sendmail, the command line option
6091 and sets the incoming protocol and host name (for trusted callers). The
6092 host name and its colon can be omitted when only the protocol is to be set.
6093 Note the Exim already has two private options, -pd and -ps, that refer to
6094 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6095 "s", but I don't think that's a major issue.
6097 10. A number of refactoring changes to the code, none of which should affect
6100 (a) The number of logging options was getting close to filling up the
6101 32-bit word that was used as a bit map. I have split them into two classes:
6102 those that are passed in the argument to log_write(), and those that are
6103 only ever tested independently outside of that function. These are now in
6104 separate 32-bit words, so there is plenty of room for expansion again.
6105 There is no change in the user interface or the logging behaviour.
6107 (b) When building, for example, log lines, the code previously used a
6108 macro that called string_cat() twice, in order to add two strings. This is
6109 not really sufficiently general. Furthermore, there was one instance where
6110 it was actually wrong because one of the argument was used twice, and in
6111 one call a function was used. (As it happened, calling the function twice
6112 did not affect the overall behaviour.) The macro has been replaced by a
6113 function that can join an arbitrary number of extra strings onto a growing
6116 (c) The code for expansion conditions now uses a table and a binary chop
6117 instead of a serial search (which was left over from when there were very
6118 few conditions). Also, it now recognizes conditions like "pam" even when
6119 the relevant support is not compiled in: a suitably worded error message is
6120 given if an attempt is made to use such a condition.
6122 11. Added ${time_interval:xxxxx}.
6124 12. A bug was causing one of the ddress fields not to be passed back correctly
6125 from remote delivery subprocesses. The field in question was not being
6126 subsequently used, so this caused to problems in practice.
6128 13. Added new log selectors queue_time and deliver_time.
6130 14. Might have fixed a bug in maildirsizefile handling that threw up
6131 "unexpected character" debug warnings, and recalculated the data
6132 unnecessarily. In any case, I expanded the warning message to give more
6135 15. Added the message "Restricted characters in address" to the statements in
6136 the default ACL that block characters like @ and % in local parts.
6138 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6139 Three changes have been made:
6141 (a) There was a serious bug; a negative response to MAIL caused the whole
6142 recipient domain to be cached as invalid, thereby blocking all messages
6143 to all local parts at the same domain, from all senders. This bug has
6144 been fixed. The domain is no longer cached after a negative response to
6145 MAIL if the sender used is not empty.
6147 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6150 (c) A new callout option, "use_sender" has been added for people who want
6151 the modified behaviour.
6157 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6160 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6161 indeed breaks things for older releases.
6163 3. Added additional logging to the case where there is a problem reading data
6164 from a filter that is running in a subprocess using a pipe, in order to
6165 try to track down a specific problem.
6167 4. Testing facility fudge: when running in the test harness and attempting
6168 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6169 getting "No route to host". Convert this to a timeout.
6171 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6174 6. Some OS don't have socklen_t but use size_t instead. This affects the
6175 fifth argument of getsockopt() amongst other things. This is now
6176 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6177 can be set for individual OS. I have set it for SunOS5, OSF1, and
6178 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6179 some earlier ones do not.
6181 7. Change 4.30/15 was not doing the test caselessly.
6183 8. The standard form for an IPv6 address literal was being rejected by address
6184 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6185 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6186 this, as well as the form without the "IPv6" on the front (but only when
6187 address literals are enabled, of course).
6189 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6191 10. Exim crashed if a message with an empty sender address specified by -f
6192 encountered a router with an errors_to setting. This could be provoked only
6193 by a command such as
6197 where an empty string was supplied; "<>" did not hit this bug.
6199 11. Installed PCRE release 4.5.
6201 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6202 remained set. It is now erased.
6204 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6205 times from message ids (which are base 36 rather than the normal 62).
6207 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6208 were being counted as actual protocol errors, and logged if the log
6209 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6210 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6211 and DATA following a set of rejected RCPTs do not count as protocol errors.
6212 In other words, Exim assumes they were pipelined, though this may not
6213 actually be the case. Of course, in all cases the client gets an
6214 appropriate error code.
6216 15. If a lookup fails in an ACL condition, a message about the failure may
6217 be available; it is used if testing the ACL cannot continue, because most
6218 such messages specify what the cause of the deferral is. However, some
6219 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6220 that caused an old message to be retained and used if a later statement
6221 caused a defer, replacing the real cause of the deferral.
6223 16. If an IP address had so many PTR records that the DNS lookup buffer
6224 was not large enough to hold them, Exim could crash while trying to process
6225 the truncated data. It now detects and logs this case.
6227 17. Further to 4.21/58, another change has been made: if (and only if) the
6228 first line of a message (the first header line) ends with CRLF, a bare LF
6229 in a subsequent header line has a space inserted after it, so as not to
6230 terminate the header.
6232 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6233 unnecessarily, used atoi() instead of strtol(), and didn't check the
6234 termination when getting file sizes from file names by regex.
6236 19. Completely re-implemented the support for maildirsize files, in the light
6237 of a number of problems with the previous contributed implementation
6238 (4.30/29). In particular:
6240 . If the quota is zero, the maildirsize file is maintained, but no quota is
6243 . If the maildir directory does not exist, it is created before any attempt
6244 to write a maildirsize file.
6246 . The quota value in the file is just a cache; if the quota is changed in
6247 the transport, the new value overrides.
6249 . A regular expression is available for excluding directories from the
6252 20. The autoreply transport checks the characters in options that define the
6253 message's headers; it allows continued headers, but it was checking with
6254 isspace() after an embedded newline instead of explicitly looking for a
6257 21. If all the "regular" hosts to which an address was routed had passed their
6258 expiry times, and had not reached their retry times, the address was
6259 bounced, even if fallback hosts were defined. Now Exim should go on to try
6262 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6263 equivalent code in the SMTP transport. Some hosts send humungous responses
6264 to HELO/EHLO, more than 1024 it seems.
6266 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6267 gives compiler warnings in some environments. I've now done it "properly",
6270 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6271 (because of problems with the built-in one) was declared to return uschar *
6272 instead of char *, causing compiler failure.
6274 25. Fixed a file descriptor leak when processing alias/forward files.
6276 26. Fixed a minor format string issue in dbfn.c.
6278 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6280 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6281 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6284 29. When a sender address is verified, it is cached, to save repeating the test
6285 when there is more than one recipient in a message. However, when the
6286 verification involves a callout, it is possible for different callout
6287 options to be set for different recipients. It is too complicated to keep
6288 track of this in the cache, so now Exim always runs a verification when a
6289 callout is required, relying on the callout cache for the optimization.
6290 The overhead is duplication of the address routing, but this should not be
6293 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6294 to be verified with callout=postmaster, and the main callout worked but the
6295 postmaster check failed, the verification correctly failed. However, if a
6296 subsequent RCPT command asked for sender verification *without* the
6297 postmaster check, incorrect caching caused this verification also to fail,
6300 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6301 it was not caching the DNS options (qualify_single, search_parents) that
6302 were used when the lookup failed. A subsequent lookup with different
6303 options therefore always gave the same answer, though there were cases
6304 where it should not have. (Example: a "domains = !$mx_any" option on a
6305 dnslookup router: the "domains" option is always processed without any
6306 widening, but the router might have qualify_single set.) Now Exim uses the
6307 cached value only when the same options are set.
6309 32. Added John Jetmore's "exipick" utility to the distribution.
6311 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6312 than a timeout (e.g. a certificate is required, and is not provided), an
6313 Exim server now closes the connection immediately. Previously it waited for
6314 the client to close - but if the client is SSL, it seems that they each
6315 wait for each other, leading to a delay before one of them times out.
6317 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6318 maintained 0.8.x compatibility because I don't think many are using it, and
6319 it is clearly obsolete.
6321 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6324 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6325 client certificate was expired. A simple patch fixes this, though I don't
6326 understand the full logic of why the verify callback is called multiple
6329 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6330 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6331 which causes problems with some clients (such as the Certicom SSL Plus
6332 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6333 disables the coutermeasure allowing Eudora to connect."
6335 38. Exim was not checking that a write() to a log file succeeded. This could
6336 lead to Bad Things if a log got too big, in particular if it hit a file
6337 size limit. Exim now panics and dies if it cannot write to a log file, just
6338 as it does if it cannot open a log file.
6340 39. Modified OS/Makefile-Linux so that it now contains
6342 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6344 The two -D definitions ensure that Exim is compiled with large file
6345 support, which makes it possible to handle log files that are bigger than
6348 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6349 instance) a domain was checked against a named list that involved a lookup,
6350 causing $domain_data to be set, then another domain was checked against the
6351 same list, then the first domain was re-checked, the value of $domain_data
6352 after the final check could be wrong. In particular, if the second check
6353 failed, it could be set empty. This bug probably also applied to
6356 41. The strip_trailing_dot option was not being applied to the address given
6357 with the -f command-line option.
6359 42. The code for reading a message's header from the spool was incrementing
6360 $received_count, but never initializing it. This meant that the value was
6361 incorrect (doubled) while delivering a message in the same process in which
6362 it was received. In the most common configuration of Exim, this never
6363 happens - a fresh exec is done - but it can happen when
6364 deliver_drop_privilege is set.
6366 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6367 it now includes up to 150 characters of the unexpected data in the log
6370 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6371 and building data strings. The size of both of these buffers was 10 000
6372 bytes - far larger than anybody would *ever* want, thought I. Needless to
6373 say, somebody hit the limit. I have increased the maximum line length to
6374 20 000 and the maximum data length of concatenated lines to 100 000. I have
6375 also fixed two bugs, because there was no checking on these buffers. Tsk,
6376 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6377 buffer is too small.
6379 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6380 lsearch lookups. Now it does.
6382 46. When parsing a route_list item in a manualroute router, a fixed-length
6383 buffer was used for the list of hosts. I made this 1024 bytes long,
6384 thinking that nobody would ever have a list of hosts that long. Wrong.
6385 Somebody had a whole pile of complicated expansion conditions, and the
6386 string was silently truncated, leading to an expansion error. It turns out
6387 that it is easier to change to an unlimited length (owing to other changes
6388 that have happened since this code was originally written) than to build
6389 structure for giving a limitation error. The length of the item that
6390 expands into the list of hosts is now unlimited.
6392 47. The lsearch lookup could not handle data where the length of text line was
6393 more than 4095 characters. Such lines were truncated, leading to shortened
6394 data being returned. It should now handle lines of any length.
6396 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6397 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6400 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6401 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6402 no longer happens in comment lines. A list of replacements is now placed
6403 at the head of all of the source files, except those whose only change is
6404 to replace PERL_COMMAND in the very first #! line.
6406 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6407 messages on the queue, with a bottom-up merge sort, using code contributed
6408 by Michael Haardt. This should make operations like -bp somewhat faster on
6409 large queues. It won't affect queue runners, except when queue_run_in_order
6412 51. Installed eximstats 1.31 in the distribution.
6414 52. Added support for SRV lookups to the dnslookup router.
6416 53. If an ACL referred to $message_body or $message_body_end, the value was not
6417 reset for any messages that followed in the same SMTP session.
6419 54. The store-handling optimization for building very long strings was not
6420 differentiating between the different store pools. I don't think this
6421 actually made any difference in practice, but I've tidied it.
6423 55. While running the routers to verify a sender address, $sender_address
6424 was still set to the sender address. This is wrong, because when routing to
6425 send a bounce to the sender, it would be empty. Therefore, I have changed
6426 it so that, while verifying a sender address, $sender_address is set to <>.
6427 (There is no change to what happens when verifying a recipient address.)
6429 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6430 target A or AAAA records (if not already returned) without resetting the
6431 qualify_single or search_parents options of the DNS resolver. These are
6432 inappropriate in this case because the targets of MX and SRV records must
6433 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6434 target that, when qualified, matched something in the local domain. These
6435 two options are now turned off when doing these lookups.
6437 57. It seems that at least some releases of Reiserfs (which does not have the
6438 concept of a fixed number of inodes) returns zero and not -1 for the
6439 number of available inodes. This interacted badly with check_spool_inodes,
6440 which assumed that -1 was the "no such thing" setting. What I have done is
6441 to check that the total number of inodes is greater than zero before doing
6442 the test of how many are available.
6444 58. When a "warn" ACL statement has a log_message modifier, the message is
6445 remembered, and not repeated. This is to avoid a lot of repetition when a
6446 message has many recipients that cause the same warning to be written.
6447 However, Exim was preserving the list of already written lines for an
6448 entire SMTP session, which doesn't seem right. The memory is now reset if a
6449 new message is started.
6451 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6452 debugging output unless log_rewrite was also set.
6454 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6455 of (int)(handle) when we know that handle contains (void *)(-1).
6457 61. The Exim daemon panic-logs an error return when it closes the incoming
6458 connection. However "connection reset by peer" seems to be common, and
6459 isn't really an error worthy of noting specially, so that particular error
6462 62. When Exim is trying to find all the local interfaces, it used to panic and
6463 die if the ioctl to get the interface flags failed. However, it seems that
6464 on at least one OS (Solaris 9) it is possible to have an interface that is
6465 included in the list of interfaces, but for which you get a failure error
6466 for this call. This happens when the interface is not "plumbed" into a
6467 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6468 failure of the "get flags" call assumes that the interface is down.
6470 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6471 makes life easier for people who are doing arithmetic on fields extracted
6472 from dates, where you often get leading zeros that should not be
6473 interpreted as octal.
6475 64. Added qualify_domain to the redirect router, to override the global
6478 65. If a pathologically long header line contained very many addresses (the
6479 report of this problem mentioned 10 000) and each of them was rewritten,
6480 Exim could use up a very large amount of memory. (It kept on making new
6481 copies of the header line as it rewrote, and never released the old ones.)
6482 At the expense of a bit more processing, the header rewriting function has
6483 been changed so that it no longer eats memory in this way.
6485 66. The generation of the Received: header has been moved from the time that a
6486 message starts to be received, to the time that it finishes. The timestamp
6487 in the Received: header should now be very close to that of the <= log
6488 line. There are two side-effects of this change:
6490 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6491 logged header lines no longer include the local Received: line, because
6492 it has not yet been created. The same applies to a copy of the message
6493 that is returned to a non-SMTP sender when a message is rejected.
6495 (b) When a filter file is tested using -bf, no additional Received: header
6496 is added to the test message. After some thought, I decided that this
6499 This change does not affect the value of $received_for. It is still set
6500 after address rewriting, but before local_scan() is called.
6502 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6504 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6505 gave an unhelpful panic error message, and a defer error. I have managed to
6506 change this behaviour so that it now rejects any supplied certificate,
6507 which seems right, as the list of acceptable certificates is empty.
6509 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6510 gave an unhelpful defer error. I have not managed to make this reject any
6511 supplied certificates, but the error message it gives is "no certificate
6512 supplied", which is not helpful.
6514 70. exigrep's output now also includes lines that are not associated with any
6515 message, but which match the given pattern. Implemented by a patch from
6516 Martin Sluka, which also tidied up the Perl a bit.
6518 71. Recipient callout verification, like sender verification, was using <> in
6519 the MAIL FROM command. This isn't really the right thing, since the actual
6520 sender may affect whether the remote host accepts the recipient or not. I
6521 have changed it to use the actual sender in the callout; this means that
6522 the cache record is now keyed on a recipient/sender pair, not just the
6523 recipient address. There doesn't seem to be a real danger of callout loops,
6524 since a callout by the remote host to check the sender would use <>.
6525 [SEE ABOVE: changed after hitting problems.]
6527 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6528 temporary errors. However, in the case of such a code being given after
6529 the end of a data transmission (i.e. after ".") Exim was failing to write
6530 a retry record for the message. (Yes, there was some broken host that was
6531 actually sending 8xx at this point.)
6533 73. An unknown lookup type in a host list could cause Exim to panic-die when
6534 the list was checked. (An example that provoked this was putting <; in the
6535 middle of a list instead of at the start.) If this happened during a DATA
6536 ACL check, a -D file could be left lying around. This kind of configuration
6537 error no longer causes Exim to die; instead it causes a defer error. The
6538 incident is still logged to the main and panic logs.
6540 74. Buglet left over from Exim 3 conversion. The message "too many messages
6541 in one connection" was written to the rejectlog but not the mainlog, except
6542 when address rewriting (yes!) was being logged.
6544 75. Added write_rejectlog option.
6546 76. When a system filter was run not as root (that is, when system_filter_user
6547 was set), the values of the $n variables were not being returned to the
6548 main process; thus, they were not subsequently available in the $sn
6551 77. Added +return_path_on_delivery log selector.
6553 78. A connection timeout was being treated differently from recipients deferred
6554 when testing hosts_max_try with a message that was older than the host's
6555 retry timeout. (The host should not be counted, thus allowing all hosts to
6556 be tried at least once before bouncing.) This may have been the cause of an
6557 occasionally reported bug whereby a message would remain on the queue
6558 longer than the retry timeout, but would be bounced if a delivery was
6559 forced. I say "may" because I never totally pinned down the problem;
6560 setting up timeout/retry tests is difficult. See also the next item.
6562 79. The ultimate address timeout was not being applied to errors that involved
6563 a combination of host plus message (for example, a timeout on a MAIL
6564 command). When an address resolved to a number of possible hosts, and they
6565 were not all tried for each delivery (e.g. because of hosts_max_try), a
6566 message could remain on the queue longer than the retry timeout.
6568 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6571 81. Fixed an obscure SMTP outgoing bug which required at least the following
6572 conditions: (a) there was another message waiting for the same server;
6573 (b) the server returned 5xx to all RCPT commands in the first message so
6574 that the message was not completed; (c) the server dropped the connection
6575 or gave a negative response to the RSET that Exim sends to abort the
6576 transaction. The observed case was a dropped connection after DATA that had
6577 been sent in pipelining mode. That is, the server had advertised PIPELINING
6578 but was not implementing it correctly. The effect of the bug was incorrect
6579 behaviour, such as trying another host, and this could lead to a crash.
6585 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6586 and daemon.c were passed as pointers to ints; they should have been
6587 pointers to socklen_t variables (which are typically unsigned ints).
6589 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6592 3. Fixed a really odd bug that affected only the testing scheme; patching a
6593 certain fixed string in the binary changed the value of another string that
6594 happened to be identical to the end of the original first string.
6596 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6597 name", it returns that address as the IP address. On some operating
6598 systems (e.g. Solaris), it also passes back the IP address string as the
6599 "host name". However, on others (e.g. Linux), it passes back an empty
6600 string. Exim wasn't checking for this, and was changing the host name to an
6601 empty string, assuming it had been canonicalized.
6603 5. Although rare, it is permitted to have more than one PTR record for a given
6604 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6605 all the names associated with an address, because they do in Solaris.
6606 However, it seems that they do not in Linux for data that comes from the
6607 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6608 I found this out when I moved to a new Linux workstation and tried to run
6609 the Exim test suite.
6611 To get round this problem I have changed the code so that it now does its
6612 own call to the DNS to look up PTR records when searching for a host name.
6613 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6614 addresses that are only in /etc/hosts are still found.
6616 This behaviour is, however, controlled by an option called host_lookup_
6617 order, which defaults to "bydns:byaddr". If people want to use the other
6618 order, or indeed, just use one or the other means of lookup, they can
6619 specify it in this variable.
6621 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6622 some operating systems, this comes back from gethostbyaddr() as an empty
6623 string, and this is what Exim used to test for. However, it seems that in
6624 other systems, "." is yielded. Exim now tests for this case too.
6626 7. The values of check_spool_space and check_log_space are now held internally
6627 as a number of kilobytes instead of an absolute number of bytes. If a
6628 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6629 kilobyte. This means that much larger values can be stored.
6631 8. Exim monitor: an attempt to get the action menu when not actually pointing
6632 at a message produces an empty menu entitled "No message selected". This
6633 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6634 no entries in it ("Shell widget menu has zero width and/or height"). So I
6635 have added a single, blank menu entry in this case.
6637 9. Added ${quote_local_part.
6639 10. MIME decoding is now applied to the contents of Subject: header lines when
6642 11. Now that a reference to $sender_host_address automatically causes a reverse
6643 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6644 host lookup before query-style lookups in lists that might use this
6645 variable. This has therefore been abolished, and the "net-" prefix is no
6646 longer necessary for query-style lookups.
6648 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6649 have been a typo for LFLAGS, so it has been changed.
6651 13. The install script calls Exim with "-C /dev/null" in order to find the
6652 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6653 to be output. However, since Exim outputs its version number before the
6654 error, it didn't break the script. It just looked ugly. I fixed this by
6655 always allowing "-C /dev/null" if the caller is root.
6657 14. Ignore overlarge ACL variable number when reading spool file - insurance
6658 against a later release with more variables having written the file.
6660 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6661 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6662 this, as well as the form without the "IPv6" on the front.
6664 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6665 OS/Makefile-Darwin file.
6667 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6668 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6670 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6671 variable. However, some OS use ints for inodes. I've added cast to long int
6672 to get rid of the compiler warning.
6674 19. I had forgotten to lock out "/../" in configuration file names when
6675 ALT_CONFIG_PREFIX was set.
6677 20. Routers used for verification do not need to specify transports. However,
6678 if such a router generated a host list, and callout was configured, Exim
6679 crashed, because it could not find a port number from the (non-existent)
6680 transport. It now assumes port 25 in this circumstance.
6682 21. Added the -t option to exigrep.
6684 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6685 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6686 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6687 them is set, LOOKUP_LSEARCH is forced.
6689 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6691 24. Added sender and host information to the "rejected by local_scan()" log
6692 line; previously there was no indication of these.
6694 25. Added .include_if_exists.
6696 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6697 Linux. It turns out that not all file systems support this. Apparently some
6698 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6699 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6700 not supported on the file descriptor, is now ignored when Exim is trying to
6701 sync a directory. This applies only to Linux.
6703 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6705 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6706 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6709 29. Added support for maildirsize files from supplied patch (modified a bit).
6711 30. The use of :fail: followed by an empty string could lead Exim to respond to
6712 sender verification failures with (e.g.):
6714 550 Verification failed for <xxx>
6715 550 Sender verify failed
6717 where the first response line was missing the '-' that indicates it is not
6718 the final line of the response.
6720 31. The loop for finding the name of the user that called Exim had a hardwired
6721 limit of 10; it now uses the value of finduser_retries, which is used for
6722 all other user lookups.
6724 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6727 33. Exim was neglecting to zero errno before one call of strtol() when
6728 expanding a string and expecting an integer value. On some systems this
6729 resulted in spurious "integer overflow" errors. Also, it was casting the
6730 result into an int without checking.
6732 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6733 did not work. The code looks as if it has *never* worked, though it appears
6734 to have been documented since at least release 1.62. I have made it work.
6736 35. The "timeout_DNS" error in retry rules, also documented since at least
6737 1.62, also never worked. As it isn't clear exactly what this means, and
6738 clearly it isn't a major issue, I have abolished the feature by treating it
6739 as "timeout", and writing a warning to the main and panic logs.
6741 36. The display of retry rules for -brt wasn't always showing the error code
6744 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6745 timeout_connect_A, timeout_connect_MX.
6747 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6748 to the empty sender.
6750 39. The daemon was not analysing the content of -oX till after it had closed
6751 stderr and disconnected from the controlling terminal. This meant that any
6752 syntax errors were only noted on the panic log, and the return code from
6753 the command was 0. By re-arranging the code a little, I've made the
6754 decoding happen first, so such errors now appear on stderr, and the return
6755 code is 1. However, the actual setting up of the sockets still happens in
6756 the disconnected process, so errors there are still only recorded on the
6759 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6760 connections (as happens on some IP stacks) was logged at start up time as
6761 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6762 it from "IPv6 and IPv4", which means that two separate sockets are being
6765 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6766 says whether AF_INET or AF_INET6 was passed as an argument.
6768 42. Exiwhat output was messed up when time zones were included in log
6771 43. Exiwhat now gives more information about the daemon's listening ports,
6772 and whether -tls-on-connect was used.
6774 44. The "port" option of the smtp transport is now expanded.
6776 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6777 silently ignored. Now an error message is written to the main and panic
6780 46. There's a new ACL modifier called "logwrite" which writes to a log file
6781 as soon as it is encountered.
6783 47. Added $local_user_uid and $local_user_gid at routing time.
6785 48. Exim crashed when trying to verify a sender address that was being
6788 49. Exim was recognizing only a space character after ".include". It now also
6789 recognizes a tab character.
6791 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6792 extracting the relevant information from the specification. The man page no
6793 longer contains scrambled data for the -d option, and I've added a section
6794 at the front about calling Exim under different names.
6796 51. Added "extra_headers" argument to the "mail" command in filter files.
6798 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6801 53. Installed eximstats 1.29.
6803 54. Added transport_filter_timeout as a generic transport option.
6805 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6806 Cc: header lines. This was required by RFC 822, but it not required by RFC
6809 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6810 incoming messages that did not have them. Now it does so only if the
6811 message originates locally, that is, if there is no associated remote host
6812 address. When Resent- header lines are present, this applies to the Resent-
6813 lines rather than the non-Resent- lines.
6815 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6816 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6818 58. Messages for configuration errors now include the name of the main
6819 configuration file - useful now that there may be more than one file in a
6820 list (.included file names were always shown).
6822 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6823 for those rare installations that do not start the daemon as root or run it
6824 setuid root. I've cut out the call to initgroups() if the daemon is not
6827 60. The Exim user and group can now be bound into the binary as text strings
6828 that are looked up at the start of Exim's processing.
6830 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6832 62. Added $mailstore_basename variable.
6834 63. Installed patch to sieve.c from Michael Haardt.
6836 64. When Exim failed to open the panic log after failing to open the main log,
6837 the original message it was trying to log was written to stderr and debug
6838 output, but if they were not available (the usual case in production), it
6839 was lost. Now it is written to syslog before the two lines that record the
6840 failures to open the logs.
6842 65. Users' Exim filters run in subprocesses under the user's uid. It is
6843 possible for a "deliver" command or an alias in a "personal" command to
6844 provoke an address rewrite. If logging of address rewriting is configured,
6845 this fails because the process is not running as root or exim. There may be
6846 a better way of dealing with this, but for the moment (because 4.30 needs
6847 to be released), I have disabled address rewrite logging when running a
6848 filter in a non-root, non-exim process.
6854 1. The buildconfig auxiliary program wasn't quoting the value set for
6855 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6856 not defined. This bug was masked in 4.22 by the effect that was fixed in
6859 2. Some messages that were rejected after a message id was allocated were
6860 shown as "incomplete" by exigrep. It no longer does this for messages that
6861 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6863 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6864 have allow_domain_literals set, the ID did not get logged in the <= line.
6865 Domain literals are now always recognized in Message-ID: header lines.
6867 4. The first argument for a ${extract expansion item is the key name or field
6868 number. Leading and trailing spaces in this item were not being ignored,
6869 causing some misleading effects.
6871 5. When deliver_drop_privilege was set, single queue runner processes started
6872 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6873 same command in the process it spins off) were not dropping privilege.
6875 6. When the daemon running as "exim" started a queue runner, it always
6876 re-executed Exim in the spun-off process. This is a waste of effort when
6877 deliver_drop_privilege is set. The new process now just calls the
6878 queue-runner function directly.
6884 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6887 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6888 ignored. Though the use of -oP was forcing the writing of a pid file, it
6889 was always written to the default place.
6891 3. If the message "no IP address found for host xxxx" is generated during
6892 incoming verification, it is now followed by identification of the incoming
6893 connection (so you can more easily find what provoked it).
6895 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6897 5. Added some features to "harden" Exim a bit more against certain attacks:
6899 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6900 be put in Local/Makefile. This is like the never_users runtime option,
6901 but it cannot be overridden. The default setting is "root".
6903 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6904 prefix string with which any file named in a -C command line option
6907 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6908 is retained for -C and -D only if the caller of Exim is root. Without
6909 it, the exim user may also use -C and -D and retain privilege.
6911 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6912 command line option is disabled.
6914 6. Macro names set by the -D option must start with an upper case letter, just
6915 like macro names defined in the configuration file.
6917 7. Added "dereference=" facility to LDAP.
6919 8. Two instances of the typo "uknown" in the source files are fixed.
6921 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6922 the Configure-Makefile script screwed up while processing it.
6924 10. Incorporated PCRE 4.4.
6926 11. The SMTP synchronization check was not operating right at the start of an
6927 SMTP session. For example, it could not catch a HELO sent before the client
6928 waited for the greeting. There is now a check for outstanding input at the
6929 point when the greeting is written. Because of the duplex, asynchronous
6930 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6931 way, but not yet received, when the check is performed.
6933 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6934 on TCP/IP sockets, because this apparently causes some broken clients to
6937 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6938 unchanged) from the Cygwin maintainer.
6940 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6941 was supported instead of testing for mbx. Effectively a typo.
6943 15. The spa authenticator server code was not checking that the input it
6944 received was valid base64.
6946 16. The debug output line for the "set" modifier in ACLs was not showing the
6947 name of the variable that was being set.
6949 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6951 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6952 lookup on its own. Something else was needed to trigger the lookup. For
6953 example, a match in host_lookup or the need for a host name in a host list.
6954 Now, if $sender_host_name is referenced and the host name has not yet been
6955 looked up, a lookup is performed. If the lookup fails, the variable remains
6956 empty, and $host_lookup_failed is set to "1".
6958 19. Added "eqi" as a case-independent comparison operator.
6960 20. The saslauthd authentication condition could segfault if neither service
6961 nor realm was specified.
6963 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6964 error message that was logged was misleading, and incoming SMTP
6965 connections were dropped. The message is now more accurate, and temporary
6966 errors are given to SMTP connections.
6968 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6969 (except RSET) with a 421 error, until QUIT is received. However, it was
6970 failing to send a response to QUIT.
6972 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6973 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6974 ACL is now run after the helo_try_verify_hosts code.
6976 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6977 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6978 case-independent, so other case variants are also recognized). Apparently
6979 some systems use these upper case variants.
6981 25. If more than two messages were waiting for the same host, and a transport
6982 filter was specified for the transport, Exim sent two messages over the
6983 same TCP/IP connection, and then failed with "socket operation on non-
6984 socket" when it tried to send the third.
6986 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6988 27. The extern definition of crypt16() in expand.c was not being excluded when
6989 the OS had its own crypt16() function.
6991 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6992 as a preferred synonym for return_size_limit, both as an option and as an
6995 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6997 30. Changed the default configuration ACL to relax the local part checking rule
6998 for addresses that are not in any local domains. For these addresses,
6999 slashes and pipe symbols are allowed within local parts, but the sequence
7000 /../ is explicitly forbidden.
7002 31. SPA server authentication was not clearing the challenge buffer before
7005 32. log_message in a "warn" ACL statement was writing to the reject log as
7006 well as to the main log, which contradicts the documentation and doesn't
7007 seem right (because no rejection is happening). So I have stopped it.
7009 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7010 However, I am unable to do any testing of this.
7012 34. Fixed an infelicity in the appendfile transport. When checking directories
7013 for a mailbox, to see if any needed to be created, it was accidentally
7014 using path names with one or more superfluous leading slashes; tracing
7015 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7017 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7018 contents are added to the log line that is written for every discarded
7019 recipient. (Previously a log_message setting was ignored.)
7021 36. The ${quote: operator now quotes the string if it is empty.
7023 37. The install script runs exim in order to find its version number. If for
7024 some reason other than non-existence or emptiness, which it checks, it
7025 could not run './exim', it was installing it with an empty version number,
7026 i.e. as "exim-". This error state is now caught, and the installation is
7029 38. An argument was missing from the function that creates an error message
7030 when Exim fails to connect to the socket for saslauthd authentication.
7031 This could cause Exim to crash, or give a corrupted message.
7033 39. Added isip, isip4, and isip6 to ${if conditions.
7035 40. The ACL variables $acl_xx are now saved with the message, and can be
7036 accessed later in routers, transports, and filters.
7038 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7039 strings in the file are not string-expanded.
7041 42. If a MAIL command specified a SIZE value that was too large to fit into an
7042 int variable, the check against message_size_limit failed. Such values are
7043 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7044 day this will have to be increased, but I don't think I want to be around
7045 when emails are that large.
7052 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7053 iconv() is not standard in FreeBSD.
7055 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7056 IPv6 enabled. The observed symptom was a segmentation fault on return from
7057 the function os_common_find_running_interfaces() in src/os.c.
7059 3. In the check_special_case() function in daemon.c I had used "errno" as an
7060 argument name, which causes warnings on some systems. This was basically a
7061 typo, since it was named "eno" in the comments!
7063 4. The code that waits for the clock to tick (at a resolution of some fraction
7064 of a second) so as to ensure message-id uniqueness was always waiting for
7065 at least one whole tick, when it could have waited for less. [This is
7066 almost certainly not relevant at current processor speeds, where it is
7067 unlikely to ever wait at all. But we try to future-proof.]
7069 5. The function that sleeps for a time interval that includes fractions of a
7070 second contained a race. It did not block SIGALRM between setting the
7071 timer, and suspending (a couple of lines later). If the interval was short
7072 and the sigsuspend() was delayed until after it had expired, the suspension
7073 never ended. On busy systems this could lead to processes getting stuck for
7076 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7077 process, before it forks any delivery processes. The open lookup caching
7078 mechanism meant that the open file or database connection was passed into
7079 the delivery process. The problem was that delivery processes always tidy
7080 up cached lookup data. This could cause a problem for the next delivery
7081 process started by the queue runner, because the external queue runner
7082 process does not know about the closure. So the next delivery process
7083 still has data in the lookup cache. In the case of a file lookup, there was
7084 no problem because closing a file descriptor in a subprocess doesn't affect
7085 the parent. However, if the lookup was caching a connection to a database,
7086 the connection was closed, and the second delivery process was likely to
7087 see errors such as "PGSQL: query failed: server closed the connection
7088 unexpectedly". The problem has been fixed by closing all cached lookups
7089 in a queue runner before running a delivery process.
7091 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7092 seem to have the "const" qualifier which it has on other OS. I've
7095 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7096 *of the same type* (client or server) with the same public name that an
7097 error should be diagnosed.
7099 9. When Exim looked up a host name for an IP address, but failed to find the
7100 original IP address when looking up the host name (a safety check), it
7101 output the message "<ip address> does not match any IP for NULL", which was
7102 confusing, to say the least. The bug was that the host name should have
7103 appeared instead of "NULL".
7105 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7106 user that is built into the binary, and the -C or -D options is used, root
7107 privilege is dropped before the configuration file is read. In addition,
7108 logging is switched to stderr instead of the normal log files. If the
7109 configuration then re-defines the Exim user, the unprivileged environment
7110 is probably not what is expected, so Exim logs a panic warning message (but
7113 However, if deliver_drop_privilege is set, the unprivileged state may well
7114 be exactly what is intended, so the warning has been cut out in that case,
7115 and Exim is allowed to try to write to its normal log files.
7121 1. smtp_return_error_details was not giving details for temporary sender
7122 or receiver verification errors.
7124 2. Diagnose a configuration error if two authenticators have the same public
7127 3. Exim used not to create the message log file for a message until the first
7128 delivery attempt. This could be confusing when incoming messages were held
7129 for policy or load reasons. The message log file is now created at the time
7130 the message is received, and an initial "Received" line is written to it.
7132 4. The automatically generated man page for command line options had a minor
7133 bug that caused no ill effects; however, a more serious problem was that
7134 the procedure for building the man page automatically didn't always
7135 operate. Consequently, release 4.20 contains an out-of-date version. This
7136 shouldn't happen again.
7138 5. When building Exim with embedded Perl support, the script that builds the
7139 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7140 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7142 6. The freeze_tell option was not being used for messages that were frozen on
7143 arrival, either by an ACL or by local_scan().
7145 7. Added the smtp_incomplete_transaction log selector.
7147 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7148 was accepting AUTH without a new EHLO.
7150 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7151 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7154 10. Logging of TCP/IP connections (when configured) now happens in the main
7155 daemon process instead of the child process, so that the TCP/IP connection
7156 count is more accurate (but it can never be perfect).
7158 11. The use of "drop" in a nested ACL was not being handled correctly in the
7159 outer ACL. Now, if condition failure induced by the nested "drop" causes
7160 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7161 or "require"), the connection is dropped.
7163 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7164 that yield "discard" can now be used with an "accept" or a "discard" verb,
7165 but an error is generated for any others (because I can't see a useful way
7166 to define what should happen).
7168 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7169 are now processed in the same way as lines in the Exim configuration file.
7170 In particular, continuation lines are supported.
7172 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7174 15. Added -ti meaning -t -i.
7176 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7177 domains, and warn by logging if others are found.
7179 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7180 returned from GIFCONF when Exim is trying to find the list of interfaces on
7181 a host. The code in os.c has been modified to copy each ifreq to an aligned
7182 structure in all cases.
7184 Also, in some cases, the returned ifreq's were being copied to a 'struct
7185 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7186 means the last couple of bytes of an IPv6 address could be chopped if the
7187 ifreq contained only a normal sockaddr (14 bytes storage).
7189 18. Named domain lists were not supported in the hosts_treat_as_local option.
7190 An entry such as +xxxx was not recognized, and was treated as a literal
7193 19. Ensure that header lines added by a DATA ACL are included in the reject log
7194 if the ACL subsequently rejects the message.
7196 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7197 MD5 (which is deprecated).
7199 21. When testing a filter file using -bf, Exim was writing a message when it
7200 took the sender from a "From " line in the message, but it was not doing so
7201 when it took $return_path from a Return-Path: header line. It now does.
7203 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7204 with a valid header line field name (a series of printing characters
7205 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7207 23. Changed "disc" in the source to "disk" to conform to the documentation and
7208 the book and for uniformity.
7210 24. Ignore Sendmail's -Ooption=value command line item.
7212 25. When execve() failed while trying to run a command in a pipe transport,
7213 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7214 could be confused with a return value of 69 from the command itself. This
7215 has been changed to 127, the value the shell returns if it is asked to run
7216 a non-existent command. The wording for the related log line suggests a
7217 non-existent command as the problem.
7219 26. If received_header_text expands to an empty string, do not add a Received:
7220 header line to the message. (Well, it adds a token one on the spool, but
7221 marks it "old" so that it doesn't get used or transmitted.)
7223 27. Installed eximstats 1.28 (addition of -nt option).
7225 28. There was no check for failure on the call to getsockname() in the daemon
7226 code. This can fail if there is a shortage of resources on the system, with
7227 ENOMEM, for example. A temporary error is now given on failure.
7229 29. Contrary to the C standard, it seems that in some environments, the
7230 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7231 program. Exim now does this explicitly; it affects the formatting of
7232 timestamps using strftime().
7234 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7235 complaints. I've now initialized all the variables, to avoid this.
7237 32. Header lines added by a system filter were not being "seen" during
7238 transport-time rewrites.
7240 33. The info_callback() function passed to OpenSSL is set up with type void
7241 (*)(SSL *, int, int), as described somewhere. However, when calling the
7242 function (actually a macro) that sets it up, the type void(*)() is
7243 expected. I've put in a cast to prevent warnings from picky compilers.
7245 34. If a DNS black list lookup found a CNAME record, but there were no A
7246 records associated with the domain it pointed at, Exim crashed.
7248 35. If a DNS black list lookup returned more than one A record, Exim ignored
7249 all but the first. It now scans all returned addresses if a particular IP
7250 value is being sought. In this situation, the contents of the
7251 $dnslist_value variable are a list of all the addresses, separated by a
7254 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7255 to accept a host name and all its aliases if the forward lookup for any of
7256 them yielded the IP address of the incoming connection. Now it accepts only
7257 those names whose forward lookup yields the correct IP address. Any other
7258 names are discarded. This closes a loophole whereby a rogue DNS
7259 administrator could create reverse DNS records to break through a
7260 wildcarded host restriction in an ACL.
7262 37. If a user filter or a system filter that ran in a subprocess used any of
7263 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7264 the wrong values were passed to the pipe command ($thisaddress had the
7265 value of $0, $0 had the value of $1, etc). This bug was introduced by
7266 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7268 38. Improved the line breaking for long SMTP error messages from ACLs.
7269 Previously, if there was no break point between 40 and 75 characters, Exim
7270 left the rest of the message alone. Two changes have been made: (a) I've
7271 reduced the minimum length to 35 characters; (b) if it can't find a break
7272 point between 35 and 75 characters, it looks ahead and uses the first one
7273 that it finds. This may give the occasional overlong line, but at least the
7274 remaining text gets split now.
7276 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7277 file descriptors might be low, and that setting 1000 would always raise it.
7278 It turns out that in some environments, the limit is already over 1000 and
7279 that lowering it causes trouble. So now Exim takes care not to decrease it.
7281 40. When delivering a message, the value of $return_path is set to $sender_
7282 address at the start of routing (routers may change the value). By an
7283 oversight, this default was not being set up when an address was tested by
7284 -bt or -bv, which affected the outcome if any router or filter referred to
7287 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7288 log only when "message" or "log_message" are set. However, if one of the
7289 conditions was an address verification, or a call to a nested ACL, the
7290 messages generated by the underlying test were being passed through. This
7291 no longer happens. The underlying message is available in $acl_verify_
7292 message for both "message" and "log_message" expansions, so it can be
7293 passed through if needed.
7295 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7296 new expansion $bh_ to give the encoded byte string without charset
7297 translation. Translation happens only if iconv() is available; HAVE_ICONV
7298 indicates this at build time. HEADERS_CHARSET gives the charset to
7299 translate to; headers_charset can change it in the configuration, and
7300 "headers charset" can change it in an individual filter file.
7302 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7303 that creates RFC 2047 encoded "words" labels them as that charset instead
7304 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7305 expansion operator; (ii) when Exim creates a From: line for a local
7306 message; (iii) when a header line is rewritten to include a "phrase" part.
7308 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7309 buggy, causing it to skip the first lines of messages whose message ID
7310 ended in 'D'. This would not have bitten before Exim release 4.14, because
7311 message IDs were unlikely to end in 'D' before then. The effect was to have
7312 incorrect size information for certain domains.
7314 45. #include "config.h" was missing at the start of the crypt16.c module. This
7315 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7318 46. If there was a timeout during a "random" callout check, Exim treated it as
7319 a failure of the random address, and carried on sending RSET and the real
7320 address. If the delay was just some slowness somewhere, the response to the
7321 original RCPT would be taken as a response to RSET and so on, causing
7322 mayhem of various kinds.
7324 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7325 when I implemented it. It didn't allow for the fact that some option values
7326 may legitimately be negative (e.g. size_addition), and it didn't even do
7327 the right test for positive values.
7329 48. Domain names in DNS records are case-independent. Exim always looks them up
7330 in lower case. Some resolvers return domain names in exactly the case they
7331 appear in the zone file, that is, they may contain uppercase letters. Not
7332 all resolvers do this - some return always lower case. Exim was treating a
7333 change of case by a resolver as a change of domain, similar to a widening
7334 of a domain abbreviation. This triggered its re-routing code and so it was
7335 trying to route what was effectively the same domain again. This normally
7336 caused routing to fail (because the router wouldn't handle the domain
7337 twice). Now Exim checks for this case specially, and just changes the
7338 casing of the domain that it ultimately uses when it transmits the message
7341 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7344 50. If a filter generated a file delivery with a non-absolute name (possible if
7345 no home directory exists for the router), the forbid_file option was not
7348 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7349 the existing equality matching.
7351 52. Exim was using ints instead of ino_t variables in some places where it was
7352 dealing with inode numbers.
7354 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7355 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7356 called TMPDIR, and if it finds it is different, it changes its value.
7358 54. The smtp_printf() function is now made available to local_scan() so
7359 additional output lines can be written before returning. There is also an
7360 smtp_fflush() function to enable the detection of a dropped connection.
7361 The variables smtp_input and smtp_batched_input are exported to
7364 55. Changed the default runtime configuration: the message "Unknown user"
7365 has been removed from the ACL, and instead placed on the localuser router,
7366 using the cannot_route_message feature. This means that any verification
7367 failures that generate their own messages won't get overridden. Similarly,
7368 the "Unrouteable address" message that was in the ACL for unverifiable
7369 relay addresses has also been removed.
7371 56. Added hosts_avoid_esmtp to the smtp transport.
7373 57. The exicyclog script was not checking for the esoteric option
7374 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7375 will work only if exicyclog is run under the appropriate euid.
7377 58. Following a discussion on the list, the rules by which Exim recognises line
7378 endings on incoming messages have been changed. The -dropcr and drop_cr
7379 options are now no-ops, retained only for backwards compatibility. The
7380 following line terminators are recognized: LF CRLF CR. However, special
7381 processing applies to CR:
7383 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7384 nor a local message in the state where . is a terminator.
7386 (ii) If a bare CR is encountered in a header line, an extra space is added
7387 after the line terminator so as not to end the header. The reasoning
7388 behind this is that bare CRs in header lines are most likely either
7389 to be mistakes, or people trying to play silly games.
7391 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7392 was being incorrectly given as 18 bytes larger than it should have been.
7393 This is a VOB (very old bug).
7395 60. This may never have affected anything current, but just in case it has:
7396 When the local host is found other than at the start of a list of hosts,
7397 the local host, those with the same MX, and any that follow, are discarded.
7398 When the list in question was part of a longer list of hosts, the following
7399 hosts (not currently being processed) were also being discarded. This no
7400 longer happens. I'm not sure if this situation could ever has previously
7403 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7406 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7407 header. This is something that is recommended in a new Internet Draft, and
7408 is something that is documented as being done by Sendmail. There are two
7409 possible values. For messages generated by the autoreply transport, Exim
7412 Auto-Submitted: auto-replied
7414 whereas for all other generated messages (e.g. bounces) it adds
7416 Auto-Submitted: auto-generated
7418 63. The "personal" condition in filters now includes a test for the
7419 Auto-Submitted: header. If it contains the string "auto-" the message it
7420 not considered personal.
7422 64. Added rcpt_include_affixes as a generic transport option.
7424 65. Added queue_only_override (default true).
7426 66. Added the syslog_duplication option.
7428 67. If what should have been the first header line of a message consisted of
7429 a space followed by a colon, Exim was mis-interpreting it as a header line.
7430 It isn't of course - it is syntactically invalid and should therefore be
7431 treated as the start of the message body. The misbehaviour could have
7432 caused a number of strange effects, including loss of data in subsequent
7433 header lines, and spool format errors.
7435 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7436 client host had authenticated. This control can now be exercised by an ACL
7437 for more flexibility.
7439 69. By default, callouts do not happen when testing with -bh. There is now a
7440 variant, -bhc, which does actually run the callout code, including
7441 consulting and updating the callout cache.
7443 70. Added support for saslauthd authentication, courtesy of Alexander
7446 71. If statvfs() failed on the spool or log directories while checking their
7447 size for availability, Exim confusingly gave the error "space shortage".
7448 Furthermore, in debugging mode it crashed with a floating point exception.
7449 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7450 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7451 serious problem, Exim now writes to the main and panic logs when this
7452 happens, with details of the failure. It then refuses to accept the
7453 incoming message, giving the message "spool directory problem" or "log
7454 directory problem" with a 421 code for SMTP messages.
7456 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7457 0, 1, and 2 exist, because some OS complain for execs without them (see
7458 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7459 descriptors. However, the code omitted to check that the open succeeded,
7460 causing mysterious errors if for some reason the permissions on /dev/null
7461 got screwed. Now Exim writes a message to the main and panic logs, and
7462 bombs out if it can't open /dev/null.
7464 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7465 interact so that it is all more flexible. It is supposed to remain
7466 backwards compatible. Also added extra_local_interfaces.
7468 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7469 to bomb out with an assertion failure - to the client this appears as a
7470 connection drop. This problem occurs in the part of the code that was taken
7471 from the Samba project. Fortunately, the assertion is in a very simple
7472 function, so I have fixed this by reproducing the function inline in the
7473 one place where it is called, and arranging for authentication to fail
7474 instead of killing the process with assert().
7476 75. The SPA client code was not working when the server requested OEM rather
7477 than Unicode encoding.
7479 76. Added code to make require_files with a specific uid setting more usable in
7480 the case where statting the file as root fails - usually a non-root-mounted
7481 NFS file system. When this happens and the failure is EACCES, Exim now
7482 forks a subprocess and does the per-uid checking as the relevant uid.
7484 77. Added process_log_path.
7486 78. If log_file_path was not explicitly set, a setting of check_log_space or
7487 check_log_inodes was ignored.
7489 79. If a space check for the spool or log partitions fails, the incident is now
7490 logged. Of course, in the latter case the data may get lost...
7492 80. Added the %p formatting code to string_format() so that it can be used to
7493 print addresses in debug_print(). Adjusted all the address printing in the
7494 debugging in store.c to use %p rather than %d.
7496 81. There was a concern that a line of code in smtp_in.c could overflow a
7497 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7498 initially expressed, the concern was not well-founded, because trailing
7499 spaces are removed early. However, if the trailing spaces were followed by
7500 a NULL, they did not get removed, so the overflow was possible. Two fixes
7503 (a) I re-wrote the offending code in a cleaner fashion.
7504 (b) If an incoming SMTP command contains a NULL character, it is rejected
7507 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7508 runs initgroups(), so that if the Exim user is in any additional groups,
7509 they will be used during message reception.
7515 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob