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.
154 JH/01 Remove code calling the customisable local_scan function, unless a new
155 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
157 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
158 non-signal-safe functions being used.
160 JH/03 Bug 2269: When presented with a received message having a stupidly large
161 number of DKIM-Signature headers, disable DKIM verification to avoid
162 a resource-consumption attack. The limit is set at twenty.
164 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
165 report of oldest_pass in ${authres } in consequence, and separate out
166 some descriptions of reasons for verification fail.
168 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
169 files in the spool were present and unlocked. A queue-runner could spot
170 them, resulting in a duplicate delivery. Fix that by doing the unlock
171 after the unlink. Investigation by Tim Stewart. Take the opportunity to
172 add more error-checking on spoolfile handling while that code is being
175 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
176 No known attacks, no CVE, this is defensive hardening.
178 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
179 a queue-runner could start a delivery while other operations were ongoing.
180 Cutthrough delivery was a common victim, resulting in duplicate delivery.
181 Found and investigated by Tim Stewart. Fix by using the open message data
182 file handle rather than opening another, and not locally closing it (which
183 releases a lock) for that case, while creating the temporary .eml format
184 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
186 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
187 $sender_verify_failure/$recipient_verify_failure to "random".
189 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
192 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
193 Previously this would segfault.
195 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
198 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
199 like zero, since the resolver should be doing this for us, But we need one
200 as a CNAME but no MX presence gets the CNAME returned; we need to check
201 that doesn't point to an MX to declare it "no MX returned" rather than
202 "error, loop". A new main option is added so the older capability of
203 following some limited number of chain links is maintained.
205 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
207 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
208 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
209 now-deprecated earlier definitions used only specified the range up to TLS
210 1.2 (in the older-version library docs).
212 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
214 JH/15 Rework TLS client-side context management. Stop using a global, and
215 explicitly pass a context around. This enables future use of TLS for
216 connections to service-daemons (eg. malware scanning) while a client smtp
217 connection is using TLS; with cutthrough connections this is quite likely.
219 JH/16 Fix ARC verification to do AS checks in reverse order.
221 JH/17 Support a "tls" option on the ${readsocket } expansion item.
223 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
224 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
225 Previously the "utf8" would be re-prepended for every additional message.
227 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
228 Previously thery were accepted, resulting in issues when attempting to
229 forward messages to a non-supporting MTA.
231 PP/02 Let -n work with printing macros too, not just options.
233 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
234 one parent address was copied, and bogus data was used at delivery-logging
235 time. Either a crash (after delivery) or bogus log data could result.
236 Discovery and analysis by Tim Stewart.
238 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
239 Previously if the string ended mid-character, we did not insert the
240 promised '?' replacement.
242 PP/04 Documentation: current string operators work on bytes, not codepoints.
244 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
245 should pack well giving a smaller memory footprint so better caching and
246 therefore performance. Group the declarations where this can't be done so
247 that the byte-sized flag variables are not interspersed among pointer
248 variables, giving a better chance of good packing by the compiler.
250 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
251 non-null, to avoid issues with sites running BATV. Previously reports were
252 sent with an empty envelope sender so looked like bounces.
254 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
255 The ignore_error flag wasn't being returned from the filter subprocess so
256 was not set for later routers. Investigation and fix by Matthias Kurz.
258 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
259 and a msg:complete for the whole, when a message is manually removed using
260 -Mrm. Developement by Matthias Kurz, hacked on by JH.
262 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
263 a "Gnu special" function, asprintf() in the DB utility binary builds; I
264 hope that is portable enough.
266 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
267 requiring a known-CA anchor certificate; make it now rely entirely on the
268 TLSA as an anchor. Checking the name on the leaf cert against the name
269 on the A-record for the host is still done for TA (but not for EE mode).
271 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
272 included in delivery lines for non-proxied connections, when compiled with
273 SUPPORT_SOCKS and running with proxy logging enabled.
275 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
276 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
277 move the existing event to fire before the normal logging of message
278 failure so that custom logging is bracketed by normal logging.
280 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
281 msg:fail:internal event. Developement by Matthias Kurz.
283 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
284 far too small for todays use of crypto signatures stored there. Go all
285 the way to the max DNS message size of 64kB, even though this might be
286 overmuch for IOT constrained device use.
288 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
289 copy a string over itself. The library routine is documented as not
290 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
292 JH/32 For main options check_spool_space and check_inode_space, where the
293 platform supports 64b integers, support more than the previous 2^31 kB
294 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
295 the previous G, M, k.
297 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
298 $authenticated_fail_id variable on authentication failure. Previously
301 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
302 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
303 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
304 GNUTLS_SEC_PARAM_MEDIUM.
306 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
307 side. Previously we would continue as if no SNI had been received.
309 JH/36 Harden the handling of string-lists. When a list consisted of a sole
310 "<" character, which should be a list-separator specification, we walked
311 off past the nul-terimation.
313 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
314 causes) even when the retry time is not yet met. Previously they were
315 not, meaning that when (say) an account was over-quota and temp-rejecting,
316 and multiple senders' messages were queued, only one sender would get
317 notified on each configured delay_warning cycle.
319 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
321 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
322 cipher-suites, an error can be left on the stack even for a succeeding
323 accept; this results in impossible error messages when a later operation
326 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
327 return error codes indicating retry. Under TLS1.3 this becomes required.
329 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
330 it only wrote the new authenticators, resulting in a lack of tracking of
331 peer changes of ESMTP extensions until the next cache flush.
333 JH/41 Fix the loop reading a message header line to check for integer overflow,
334 and more-often against header_maxsize. Previously a crafted message could
335 induce a crash of the recive process; now the message is cleanly rejected.
337 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
338 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
344 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
345 When redis_servers is set to a list of > 1 element, and the Redis servers
346 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
347 case of MOVED into a DEFER case instead, thus moving the query onto the
348 next server in the list. For a cluster of N elements, all N servers must
349 be defined in redis_servers.
351 GF/02 Catch and remove uninitialized value warning in exiqsumm
352 Check for existence of @ARGV before looking at $ARGV[0]
354 JH/01 Replace the store_release() internal interface with store_newblock(),
355 which internalises the check required to safely use the old one, plus
356 the allocate and data copy operations duplicated in both (!) of the
357 extant use locations.
359 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
360 modifier. This matches the restriction on the commandline.
362 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
363 Previously only the last row was returned.
365 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
366 we assumed that tags in the header were well-formed, and parsed the
367 element content after inspecting only the first char of the tag.
368 Assumptions at that stage could crash the receive process on malformed
371 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
372 While running the DKIM ACL we operate on the Permanent memory pool so that
373 variables created with "set" persist to the DATA ACL. Also (at any time)
374 DNS lookups that fail create cache records using the Permanent pool. But
375 expansions release any allocations made on the current pool - so a dnsdb
376 lookup expansion done in the DKIM ACL releases the memory used for the
377 DNS negative-cache, and bad things result. Solution is to switch to the
378 Main pool for expansions.
379 While we're in that code, add checks on the DNS cache during store_reset,
380 active in the testsuite.
381 Problem spotted, and debugging aided, by Wolfgang Breyha.
383 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
384 When none of the hosts presented to a transport match an already-open
385 connection, close it and proceed with the list. Previously we would
386 queue the message. Spotted by Lena with Yahoo, probably involving
389 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
390 Previously a spurious "250 OK id=" response was appended to the proper
393 JH/08 The "support for" informational output now, which built with Content
394 Scanning support, has a line for the malware scanner interfaces compiled
395 in. Interface can be individually included or not at build time.
397 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
398 by the template makefile "src/EDITME". The "STREAM" support for an older
399 ClamAV interface method is removed.
401 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
402 rows affected is given instead).
404 JH/11 The runtime Berkeley DB library version is now additionally output by
405 "exim -d -bV". Previously only the compile-time version was shown.
407 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
408 SMTP connection. Previously, when one had more receipients than the
409 first, an abortive onward connection was made. Move to full support for
410 multiple onward connections in sequence, handling cutthrough connection
411 for all multi-message initiating connections.
413 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
414 routers. Previously, a multi-recipient message would fail to match the
415 onward-connection opened for the first recipient, and cause its closure.
417 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
418 a timeout on read on a GnuTLS initiating connection, resulting in the
419 initiating connection being dropped. This mattered most when the callout
420 was marked defer_ok. Fix to keep the two timeout-detection methods
423 JH/15 Relax results from ACL control request to enable cutthrough, in
424 unsupported situations, from error to silently (except under debug)
425 ignoring. This covers use with PRDR, frozen messages, queue-only and
428 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
430 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
431 metadata, resulting in a crash in free().
433 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
434 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
435 Broken also in d185889f4, with init system revamp.
437 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
438 Previously we abruptly closed the connection after reading a malware-
439 found indication; now we go on to read the "scan ok" response line,
442 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
443 ACL. Previously, a crash would result.
445 JH/19 Speed up macro lookups during configuration file read, by skipping non-
446 macro text after a replacement (previously it was only once per line) and
447 by skipping builtin macros when searching for an uppercase lead character.
449 JH/20 DANE support moved from Experimental to mainline. The Makefile control
450 for the build is renamed.
452 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
453 was allocated for every new TLS startup, meaning one per message. Fix
454 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
456 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
457 reported the original. Fix to report (as far as possible) the ACL
458 result replacing the original.
460 JH/23 Fix memory leak during multi-message connections using STARTTLS under
461 OpenSSL. Certificate information is loaded for every new TLS startup,
462 and the resources needed to be freed.
464 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
466 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
469 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
470 DATA response info to the (existing) per-recipient response info for
471 the "C=" log element. It can have useful tracking info from the
472 destination system. Patch from Simon Arlott.
474 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
475 length value. Previously this would segfault.
477 HS/02 Support Avast multiline protoocol, this allows passing flags to
478 newer versions of the scanner.
480 JH/28 Ensure that variables possibly set during message acceptance are marked
481 dead before release of memory in the daemon loop. This stops complaints
482 about them when the debug_store option is enabled. Discovered specifically
483 for sender_rate_period, but applies to a whole set of variables.
484 Do the same for the queue-runner and queue-list loops, for variables set
485 from spool message files. Do the same for the SMTP per-message loop, for
486 certain variables indirectly set in ACL operations.
488 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
489 as a multi-recipient message from a mailinglist manager). The coding had
490 an arbitrary cutoff number of characters while checking for more input;
491 enforced by writing a NUL into the buffer. This corrupted long / fast
492 input. The problem was exposed more widely when more pipelineing of SMTP
493 responses was introduced, and one Exim system was feeding another.
494 The symptom is log complaints of SMTP syntax error (NUL chars) on the
495 receiving system, and refused recipients seen by the sending system
496 (propating to people being dropped from mailing lists).
497 Discovered and pinpointed by David Carter.
499 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
500 replaced by the ${authresults } expansion.
502 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
504 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
505 allows proper process termination in container environments.
507 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
508 Previously the "final dot" had a newline after it; ensure it is CR,LF.
510 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
511 and "err_perm", deprecated since 4.83 when the RFC-defined words
512 "temperror" and "permerror" were introduced.
514 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
515 transport-filters or DKIM-signing. The restriction was lost in the
516 consolidation of verify-callout and delivery SMTP handling.
517 Extend the restriction to also cover ARC-signing.
519 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
520 in defer=pass mode supply a 450 to the initiator. Previously the message
523 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
524 tls_require_ciphers is used as before.
526 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
527 "pass_unscanned". Only tmpfails from the scanner are written to
528 the paniclog, as they may require admin intervention (permission
529 denied, license issues). Other scanner errors (like decompression
530 bombs) do not cause a paniclog entry.
532 JH/36 Fix reinitialisation of DKIM logging variable between messages.
533 Previously it was possible to log spurious information in receive log
536 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
537 triggered odd behaviour from Outlook Express clients.
539 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
542 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
543 since the IETF WG has not yet settled on that versus the original
544 "bare" representation.
546 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
547 Previously the millisecond value corrupted the output.
548 Fix also for syslog_pid=no and log_selector +pid, for which the pid
549 corrupted the output.
555 JH/01 Rework error string handling in TLS interface so that the caller in
556 more cases is responsible for logging. This permits library-sourced
557 string to be attached to addresses during delivery, and collapses
558 pairs of long lines into single ones.
560 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
561 during configuration. Wildcards are allowed and expanded.
563 JH/02 Rework error string handling in DKIM to pass more info back to callers.
564 This permits better logging.
566 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
567 do not close it down and have the child transport start it up again on
568 the passed-on TCP connection. Instead, proxy the child (and any
569 subsequent ones) for TLS via a unix-domain socket channel. Logging is
570 affected: the continued delivery log lines do not have any DNSSEC, TLS
571 Certificate or OCSP information. TLS cipher information is still logged.
573 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
574 identical IP addresses on different listening ports. Will also affect
577 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
578 add noisy ifdef guards to special-case this sillyness.
579 Patch from Bernd Kuhls.
581 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
582 than 255 are no longer allowed.
584 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
585 Disable the session-cache too, which might reduce our load. Since we
586 currrectly use a new context for every connection, both as server and
587 client, there is no benefit for these.
588 GnuTLS appears to not support tickets server-side by default (we don't
589 call gnutls_session_ticket_enable_server()) but client side is enabled
590 by default on recent versions (3.1.3 +) unless the PFS priority string
593 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
594 <https://reproducible-builds.org/specs/source-date-epoch/>.
596 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
597 the check for any unsuccessful recipients did not notice the limit, and
598 erroneously found still-pending ones.
600 JH/08 Pipeline CHUNKING command and data together, on kernels that support
601 MSG_MORE. Only in-clear (not on TLS connections).
603 JH/09 Avoid using a temporary file during transport using dkim. Unless a
604 transport-filter is involved we can buffer the headers in memory for
605 creating the signature, and read the spool data file once for the
606 signature and again for transmission.
608 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
609 4.77 as the kernel support then wasn't solid, having issues in 64bit
610 mode. Now, it's been long enough. Add support for FreeBSD also.
612 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
613 case where the routing stage had gathered several addresses to send to
614 a host before calling the transport for the first, we previously failed
615 to close down TLS in the old transport process before passing the TCP
616 connection to the new process. The new one sent a STARTTLS command
617 which naturally failed, giving a failed delivery and bloating the retry
618 database. Investigation and fix prototype from Wolfgang Breyha.
620 JH/12 Fix check on SMTP command input synchronisation. Previously there were
621 false-negatives in the check that the sender had not preempted a response
622 or prompt from Exim (running as a server), due to that code's lack of
623 awareness of the SMTP input buffering.
625 PP/04 Add commandline_checks_require_admin option.
626 Exim drops privileges sanely, various checks such as -be aren't a
627 security problem, as long as you trust local users with access to their
628 own account. When invoked by services which pass untrusted data to
629 Exim, this might be an issue. Set this option in main configuration
630 AND make fixes to the calling application, such as using `--` to stop
633 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
634 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
635 responses to those, into a single TLS record each way (this usually means
636 a single packet). As a side issue, smtp_enforce_sync now works on TLS
639 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
640 affects you only if you're dancing at the edge of the param size limits.
641 If you are, and this message makes sense to you, then: raise the
642 configured limit or use OpenSSL 1.1. Nothing we can do for older
645 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
646 cmdline element to get the documented default one. Previously it was
649 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
650 are unsuable for non-message-specific reasons - eg. connection timeout,
651 banner-time rejection.
653 JH/16 Fix logging of delivery remote port, when specified by router, under
656 PP/06 Repair manualroute's ability to take options in any order, even if one
657 is the name of a transport.
660 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
662 JH/17 Change the list-building routines interface to use the expanding-string
663 triplet model, for better allocation and copying behaviour.
665 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
666 Previously it was constructed the first time a possibly-matching string
667 was met in the configuration file input during startup; now it is done
670 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
671 compatible one, to avoid the (poorly documented) possibility of a config
672 file in the working directory redirecting the DB files, possibly correpting
673 some existing file. CVE-2017-10140 assigned for BDB.
675 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
676 cache-hot. Previously, although the result was properly cached, the
677 initial verify call returned a defer.
679 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
680 the main verify for receipient in uncached-mode.
682 JH/22 Retire historical build files to an "unsupported" subdir. These are
683 defined as "ones for which we have no current evidence of testing".
685 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
686 if present. Previously it was ignored.
688 JH/24 Start using specified-initialisers in C structure init coding. This is
689 a C99 feature (it's 2017, so now considered safe).
691 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
692 if was a fixed-sized field and bitmask ops via macros; it is now more
695 PP/07 GitHub PR 56: Apply MariaDB build fix.
696 Patch provided by Jaroslav Škarvada.
698 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
699 during Coverity cleanups [4.87 JH/47]
700 Diagnosis and fix provided by Michael Fischer v. Mollard.
702 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
703 the right size to place the terminating semicolon on its own folded
704 line, the header hash was calculated to an incorrect value thanks to
705 the (relaxed) space the fold became.
707 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
708 and confused the parent.
710 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
711 which could crash as a result. This could lead to undeliverable messages.
713 JH/28 Logging: "next input sent too soon" now shows where input was truncated
716 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
717 matters on fast-turnover and PID-randomising systems, which were getting
718 out-of-order delivery.
720 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
721 a possibly-overlapping copy. The symptom was that "Remote host closed
722 connection in response to HELO" was logged instead of the actual 4xx
725 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
726 Previously only that bufferd was discarded, resulting in SYMTP command
729 JH/32 DKIM: when a message has multiple signatures matching an identity given
730 in dkim_verify_signers, run the dkim acl once for each. Previously only
731 one run was done. Bug 2189.
733 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
734 "panic the current process" to "deliberately defer". The panic log is
735 still written with the problem list name; the mail and reject logs now
736 get a temp-reject line for the message that was being handled, saying
737 something like "domains check lookup or other defer". The SMTP 451
738 message is still "Temporary local problem".
740 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
741 A crafted sequence of BDAT commands could result in in-use memory beeing
742 freed. CVE-2017-16943.
744 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
745 from SMTP input. Previously it was always done; now only done for DATA
746 and not BDAT commands. CVE-2017-16944.
748 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
749 to the message (such as an overlong header line). Previously this was
750 not done and we did not exit BDAT mode. Followon from the previous item
751 though a different problem.
757 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
758 than -2003 did; needs libidn2 in addition to libidn.
760 JH/02 The path option on a pipe transport is now expanded before use.
762 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
763 Patch provided by "Björn", documentation fix added too.
765 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
766 missing a wire-to-host endian conversion.
768 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
769 close after a BDAT command line could be taken as a following command,
770 giving a synch failure. Fix by only checking for synch immediately
771 before acknowledging the chunk.
773 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
774 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
775 macro. Patches provided by Josh Soref.
777 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
778 Previously we did not; the RFC seems ambiguous and VRFY is not listed
779 by IANA as a service extension. However, John Klensin suggests that we
782 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
783 the dkim code may be unix-mode line endings rather than smtp wire-format
784 CRLF, so prepend a CR to any bare LF.
786 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
787 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
789 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
790 the first were themselves being wrongly included in the feed into dkim
791 processing; with most chunk sizes in use this resulted in an incorrect
792 body hash calculated value.
794 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
795 DKIM signature block, for verification. Although advised against by
796 standards it is specifically not ruled illegal.
798 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
800 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
801 missing a body hash (the bh= tag).
803 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
804 It seems that HAProxy sends the Proxy Protocol information in clear and
805 only then does a TLS startup, so do the same.
807 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
808 TCP connections (such as for Spamd) unless the daemon successfully set
809 Fast Open mode on its listening sockets. This fixes breakage seen on
810 too-old kernels or those not configured for Fast Open, at the cost of
811 requiring both directions being enabled for TFO, and TFO never being used
812 by non-daemon-related Exim processes.
814 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
815 endings, at least on the first header line. Try to canonify any that get
816 past that check, despite the cost.
818 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
819 now limited to an arbitrary five deep, while parsing addresses with the
820 strip_excess_angle_brackets option enabled.
822 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
823 instead leave the unprompted TLS handshake in socket buffer for the
824 TLS library to consume.
826 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
828 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
830 JH/16 Drop variables when they go out of scope. Memory management drops a whole
831 region in one operation, for speed, and this leaves assigned pointers
832 dangling. Add checks run only under the testsuite which checks all
833 variables at a store-reset and panics on a dangling pointer; add code
834 explicitly nulling out all the variables discovered. Fixes one known
835 bug: a transport crash, where a dangling pointer for $sending_ip_address
836 originally assigned in a verify callout, is re-used.
838 PP/06 Drop '.' from @INC in various Perl scripts.
840 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
842 PP/08 Reduce a number of compilation warnings under clang; building with
843 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
844 should be warning-free.
846 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
848 HS/01 Fix portability problems introduced by PP/08 for platforms where
849 realloc(NULL) is not equivalent to malloc() [SunOS et al].
851 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
852 chunk. This allows us to accept broken chunked messages. We need a more
853 general solution here.
855 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
856 already-broken messages in the queue.
858 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
860 JH/19 Fix reference counting bug in routing-generated-address tracking.
866 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
867 supports it and a size is available (ie. the sending peer gave us one).
869 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
870 years of being deprecated). The replacements are the ACLs
871 acl_smtp_mime and acl_not_smtp_mime.
873 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
874 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
875 or even an in-clear connection were permitted. Now, if the host lookup
876 was dnssec and dane was requested then the host is only used if the
877 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
878 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
879 if one fails this test.
880 This means that a poorly-configured remote DNS will make it incommunicado;
881 but it protects against a DNS-interception attack on it.
883 JH/04 Bug 1810: make continued-use of an open smtp transport connection
884 non-noisy when a race steals the message being considered.
886 JH/05 If main configuration option tls_certificate is unset, generate a
887 self-signed certificate for inbound TLS connections.
889 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
890 in rewrites and routers.
892 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
893 and logged a warning sing 4.83; now they are a configuration file error.
895 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
896 (lacking @domain). Apply the same qualification processing as RCPT.
898 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
900 JH/10 Support ${sha256:} applied to a string (as well as the previous
903 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
904 a cutthrough deliver is pending, as we always want to make a connection.
905 This also avoids re-routing the message when later placing the cutthrough
906 connection after a verify cache hit.
907 Do not update it with the verify result either.
909 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
910 when routing results in more than one destination address.
912 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
913 signing (which inhibits the cutthrough capability). Previously only
914 the presence of an option was tested; now an expansion evaluating as
915 empty is permissible (obviously it should depend only on data available
916 when the cutthrough connection is made).
918 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
919 the relevant preceding SMTP command did not note the pipelining mode.
921 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
922 Previously they were not counted.
924 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
925 as one having no matching records. Previously we deferred the message
926 that needed the lookup.
928 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
929 distinguished as "(=".
931 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
932 for missing MX records. Previously it only worked for missing A records.
934 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
936 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
937 after the data-go-ahead and data-ack. Patch from Jason Betts.
939 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
940 even for a "none" policy. Patch from Tony Meyer.
942 JH/22 Fix continued use of a connection for further deliveries. If a port was
943 specified by a router, it must also match for the delivery to be
946 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
947 When one of the recipients of a message was unsuitable for the connection
948 (has no matching addresses), we lost track of needing to mark it
949 deferred. As a result mail would be lost.
951 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
953 JH/25 Decoding ACL controls is now done using a binary search; the source code
954 takes up less space and should be simpler to maintain. Merge the ACL
955 condition decode tables also, with similar effect.
957 JH/26 Fix problem with one_time used on a redirect router which returned the
958 parent address unchanged. A retry would see the parent address marked as
959 delivered, so not attempt the (identical) child. As a result mail would
962 JH/27 Fix a possible security hole, wherein a process operating with the Exim
963 UID can gain a root shell. Credit to http://www.halfdog.net/ for
964 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
967 JH/28 Enable {spool,log} filesystem space and inode checks as default.
968 Main config options check_{log,spool}_{inodes,space} are now
969 100 inodes, 10MB unless set otherwise in the configuration.
971 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
972 Previously it only applied to the main-section connection policy
975 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
977 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
978 by me. Added RFC7919 DH primes as an alternative.
980 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
981 are not in the system include path.
983 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
984 GnuTLS, when a session startup failed (eg because the client disconnected)
985 Exim did stdio operations after fclose. This was exposed by a recent
986 change which nulled out the file handle after the fclose.
988 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
989 signed directly by the cert-signing cert, rather than an intermediate
990 OCSP-signing cert. This is the model used by LetsEncrypt.
992 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
994 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
995 an incoming connection.
997 HS/02 Bug 1802: Do not half-close the connection after sending a request
1000 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1001 fallback to "prime256v1".
1003 JH/34 SECURITY: Use proper copy of DATA command in error message.
1004 Could leak key material. Remotely exploitable. CVE-2016-9963.
1010 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1011 and 3.4.4 - once the server is enabled to respond to an OCSP request
1012 it does even when not requested, resulting in a stapling non-aware
1013 client dropping the TLS connection.
1015 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1016 support variable-length bit vectors. No functional change.
1018 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1019 The I= interface field on outgoing lines is now after the H= remote
1020 host field, same as incoming lines. There is a separate
1021 outgoing_interface log selector which allows you to disable the
1024 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1025 If not running log_selector +smtp_connection the mainlog would be held
1026 open indefinitely after a "too many connections" event, including to a
1027 deleted file after a log rotate. Leave the per net connection logging
1028 leaving it open for efficiency as that will be quickly detected by the
1029 check on the next write.
1031 HS/01 Bug 1671: Fix post transport crash.
1032 Processing the wait-<transport> messages could crash the delivery
1033 process if the message IDs didn't exist for some reason. When
1034 using 'split_spool_directory=yes' the construction of the spool
1035 file name failed already, exposing the same netto behaviour.
1037 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1038 mime_regex ACL conditions.
1040 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1041 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1042 to HELO, local diagnostic string.
1044 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1045 log line to debug. Even when configured with a tls authenticator many
1046 client connections are expected to not authenticate in this way, so
1047 an authenticate fail is not an error.
1049 HS/02 Add the Exim version string to the process info. This way exiwhat
1050 gives some more detail about the running daemon.
1052 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1053 matter for fast-change records such as DNSBLs.
1055 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1056 retry record, even if constant. There may be multiple transports with
1057 different interface settings and the retry behaviour needs to be kept
1060 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1062 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1064 JH/10 Bug 840: fix log_defer_output option of pipe transport
1066 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1067 hosts_randomize. This matters a lot when combined with mua_wrapper.
1069 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1070 ${quote_pgsql:<string>} operator.
1072 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1074 JH/14 Change header file name for freeradius-client. Relevant if compiling
1075 with Radius support; from the Gentoo tree and checked under Fedora.
1077 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1079 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1080 option transport_filter, meaning no filtering.
1082 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1084 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1085 defaults to "*" (all hosts). The variable is now available when not built
1086 with TLS, default unset, mainly to enable keeping the testsuite sane.
1087 If a server certificate is not supplied (via tls_certificate) an error is
1088 logged, and clients will find TLS connections fail on startup. Presumably
1089 they will retry in-clear.
1090 Packagers of Exim are strongly encouraged to create a server certificate
1091 at installation time.
1093 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1094 with the $config_file variable.
1096 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1097 in transport context, after the attempt, and per-recipient. The latter type
1098 is per host attempted. The event data is the error message, and the errno
1099 information encodes the lookup type (A vs. MX) used for the (first) host,
1100 and the trailing two digits of the smtp 4xx response.
1102 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1103 to write to mainlog (or rejectlog, paniclog) in the window between file
1104 creation and permissions/ownership being changed. Particularly affects
1105 installations where exicyclog is run as root, rather than exim user;
1106 result is that the running daemon panics and dies.
1108 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1110 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1111 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1112 "pri" and "weight". Note that the previous implicit priority given by the
1113 list order is no longer honoured.
1115 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1116 for DKIM processing.
1118 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1119 by defining SUPPORT_SOCKS.
1121 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1122 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1123 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1124 variables are renamed to proxy_{local,external}_{address,port}.
1126 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1127 for a build by defining SUPPORT_I18N
1129 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1130 of the query string, and make ${quote_redis:} do that quoting.
1132 JH/29 Move Events support from Experimental to mainline, enabled by default
1133 and removable for a build by defining DISABLE_EVENT.
1135 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1137 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1138 cached by the daemon.
1140 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1141 by defining LOOKUP_REDIS. The libhiredis library is required.
1143 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1144 keys are given for lookup.
1146 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1147 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1148 only supported when built with TLS support. The PolarSSL SHA routines
1149 are still used when the TLS library is too old for convenient support.
1151 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1152 openssl_options), for security. OpenSSL forces this from version 1.1.0
1153 server-side so match that on older versions.
1155 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1156 allocation for $value could be released as the expansion processing
1157 concluded, but leaving the global pointer active for it.
1159 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1160 and to use the domains and local_parts ACL conditions.
1162 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1163 incorrectly not doubled on cutthrough transmission, hence seen as a
1164 body-termination at the receiving system - resulting in truncated mails.
1165 Commonly the sender saw a TCP-level error, and retransmitted the message
1166 via the normal store-and-forward channel. This could result in duplicates
1167 received - but deduplicating mailstores were liable to retain only the
1168 initial truncated version.
1170 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1172 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1174 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1175 we're in there, support oversigning also; bug 1309.
1177 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1179 HS/04 Add support for keep_environment and add_environment options.
1181 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1182 either intentional arithmetic overflow during PRNG, or testing config-
1185 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1186 delivery resulted in actual delivery. Cancel cutthrough before DATA
1189 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1190 rejects a recipient: pass the reject to the originator.
1192 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1193 Many were false-positives and ignorable, but it's worth fixing the
1196 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1197 for the new environment-manipulation done at startup. Move the routines
1198 from being local to tls.c to being global via the os.c file.
1200 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1201 an extract embedded as result-arg for a map, the first arg for extract
1202 is unavailable so we cannot tell if this is a numbered or keyed
1203 extraction. Accept either.
1209 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1212 JH/02 The smtp transport option "multi_domain" is now expanded.
1214 JH/03 The smtp transport now requests PRDR by default, if the server offers
1217 JH/04 Certificate name checking on server certificates, when exim is a client,
1218 is now done by default. The transport option tls_verify_cert_hostnames
1219 can be used to disable this per-host. The build option
1220 EXPERIMENTAL_CERTNAMES is withdrawn.
1222 JH/05 The value of the tls_verify_certificates smtp transport and main options
1223 default to the word "system" to access the system default CA bundle.
1224 For GnuTLS, only version 3.0.20 or later.
1226 JH/06 Verification of the server certificate for a TLS connection is now tried
1227 (but not required) by default. The verification status is now logged by
1228 default, for both outbound TLS and client-certificate supplying inbound
1231 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1234 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1235 Status Notification (bounce) messages are now MIME format per RFC 3464.
1236 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1237 under the control of the dsn_advertise_hosts option, and routers may
1238 have a dsn_lasthop option.
1240 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1241 default, modifiable by a malware= option. The list separator for
1242 the options can now be changed in the usual way. Bug 68.
1244 JH/10 The smtp_receive_timeout main option is now expanded before use.
1246 JH/11 The incoming_interface log option now also enables logging of the
1247 local interface on delivery outgoing connections.
1249 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1250 if the interface and destination host and port all match.
1252 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1255 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1256 Patch from Andrew Lewis.
1258 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1259 now supports optional time-restrictions, weighting, and priority
1260 modifiers per server. Patch originally by <rommer@active.by>.
1262 JH/16 The spamd_address main option now supports a mixed list of local
1263 and remote servers. Remote servers can be IPv6 addresses, and
1264 specify a port-range.
1266 JH/17 Bug 68: The spamd_address main option now supports an optional
1267 timeout value per server.
1269 JH/18 Bug 1581: Router and transport options headers_add/remove can
1270 now have the list separator specified.
1272 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1275 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1278 JH/21 Support for the A6 type of dns record is withdrawn.
1280 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1281 rather than the verbs used.
1283 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1284 from 255 to 1024 chars.
1286 JH/24 Verification callouts now attempt to use TLS by default.
1288 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1289 are generic router options now. The defaults didn't change.
1291 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1292 Original patch from Alexander Shikoff, worked over by JH.
1294 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1295 files if ZCAT_COMMAND is not executable.
1297 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1299 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1301 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1302 Normally benign, it bites when the pair was led to by a CNAME;
1303 modern usage is to not canonicalize the domain to a CNAME target
1304 (and we were inconsistent anyway for A-only vs AAAA+A).
1306 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1308 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1309 when evaluating $sender_host_dnssec.
1311 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1312 $sender_helo_dnssec variable.
1314 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1316 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1318 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1320 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1321 documented as working, but never had. Support all but $spam_report.
1323 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1324 added for tls authenticator.
1326 HS/03 Add perl_taintmode main config option
1332 TL/01 When running the test suite, the README says that variables such as
1333 no_msglog_check are global and can be placed anywhere in a specific
1334 test's script, however it was observed that placement needed to be near
1335 the beginning for it to behave that way. Changed the runtest perl
1336 script to read through the entire script once to detect and set these
1337 variables, reset to the beginning of the script, and then run through
1338 the script parsing/test process like normal.
1340 TL/02 The BSD's have an arc4random API. One of the functions to induce
1341 adding randomness was arc4random_stir(), but it has been removed in
1342 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1343 function when detected.
1345 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1346 cause callback expansion.
1348 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1349 syntax errors in an expansion can be treated as a string instead of
1350 logging or causing an error, due to the internal use of bool_lax
1351 instead of bool when processing it.
1353 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1354 server certificates when making smtp deliveries.
1356 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1358 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1360 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1362 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1363 Merged patch from Sebastian Wiedenroth.
1365 JH/05 Fix results-pipe from transport process. Several recipients, combined
1366 with certificate use, exposed issues where response data items split
1367 over buffer boundaries were not parsed properly. This eventually
1368 resulted in duplicates being sent. This issue only became common enough
1369 to notice due to the introduction of connection certificate information,
1370 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1372 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1373 size buffer was used, resulting in syntax errors when an expansion
1376 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1377 version 3.3.6 or later.
1379 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1380 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1381 both become "event_action", the variables become $event_name, $event_data
1382 and $event_defer_errno. There is a new variable $verify_mode, usable in
1383 routers, transports and related events. The tls:cert event is now also
1384 raised for inbound connections, if the main configuration event_action
1387 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1388 early OCSP support, but no stapling (appears to be less than 1.0.0).
1390 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1391 server certificate names available under the smtp transport option
1392 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1395 JH/10 Time-related extraction expansions from certificates now use the main
1396 option "timezone" setting for output formatting, and are consistent
1397 between OpenSSL and GnuTLS compilations. Bug 1541.
1399 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1400 encoded parameter in the incoming message. Bug 1558.
1402 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1403 include certificate info, eximon was claiming there were spoolfile
1406 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1408 JH/14 Log delivery-related information more consistently, using the sequence
1409 "H=<name> [<ip>]" wherever possible.
1411 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1412 are problematic for Debian distribution, omit them from the release
1415 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1417 JH/16 Fix string representation of time values on 64bit time_t architectures.
1420 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1421 output list separator was used.
1426 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1427 checkers that were complaining about end of non-void function with no
1430 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1431 This was a regression introduced in 4.83 by another bugfix.
1433 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1435 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1436 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1442 TF/01 Correctly close the server side of TLS when forking for delivery.
1444 When a message was received over SMTP with TLS, Exim failed to clear up
1445 the incoming connection properly after forking off the child process to
1446 deliver the message. In some situations the subsequent outgoing
1447 delivery connection happened to have the same fd number as the incoming
1448 connection previously had. Exim would try to use TLS and fail, logging
1449 a "Bad file descriptor" error.
1451 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1452 utilities have not been installed.
1454 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1455 temporary space as the ACL may create new global variables.
1457 TL/01 LDAP support uses per connection or global context settings, depending
1458 upon the detected version of the libraries at build time.
1460 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1461 to extract and use the src ip:port in logging and expansions as if it
1462 were a direct connection from the outside internet. PPv2 support was
1463 updated based on HAProxy spec change in May 2014.
1465 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1467 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1468 Properly escape header and check for NULL return.
1470 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1473 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1475 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1476 characters in header names, implemented as a verify condition.
1477 Contributed by Michael Fischer v. Mollard.
1479 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1480 results permerror and temperror. Previous values are deprecated but
1481 still accepted. In a future release, err_perm and err_temp will be
1482 completely removed, which will be a backward incompatibility if the
1483 ACL tests for either of these two old results. Patch contributed by
1484 user bes-internal on the mailing list.
1486 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1488 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1489 selectors, in both main and reject logs.
1491 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1494 JH/07 Add malware type "sock" for talking to simple daemon.
1496 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1498 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1499 routers/transports under cutthrough routing.
1501 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1502 numbers. Touch up "bool" conditional to keep the same definition.
1504 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1506 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1508 JH/12 Expand items in router/transport headers_add or headers_remove lists
1509 individually rather than the list as a whole. Bug 1452.
1511 Required for reasonable handling of multiple headers_ options when
1512 they may be empty; requires that headers_remove items with embedded
1513 colons must have them doubled (or the list-separator changed).
1515 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1516 view the policy declared in the DMARC record. Currently, $dmarc_status
1517 is a combined value of both the record presence and the result of the
1520 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1522 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1523 dnslookup router and the smtp transport (applying to the forward
1526 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1527 of ldap servers used for a specific lookup. Patch provided by Heiko
1530 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1531 New variable $lookup_dnssec_authenticated for observability.
1533 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1534 Patch submitted by Lars Timman.
1536 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1538 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1539 Requires trusted mode and valid format message id, aborts otherwise.
1540 Patch contributed by Heiko Schlichting.
1542 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1543 certextract with support for various fields. Bug 1358.
1545 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1546 is requested by default, modifiable by smtp transport option
1549 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1550 operate on certificate variables to give certificate fingerprints
1551 Also new ${sha256:cert_variable}.
1553 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1555 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1556 Christian Aistleitner.
1558 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1560 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1561 file. Patch from Wolfgang Breyha.
1563 JH/25 Expand the coverage of the delivery $host and $host_address to
1564 client authenticators run in verify callout. Bug 1476.
1566 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1567 align with daemon_smtp_ports. Bug 72.
1569 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1570 support and error reporting did not work properly.
1572 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1573 and is readable. Patch from Andrew Colin Kissa.
1575 TL/14 Enhance documentation of ${run expansion and how it parses the
1576 commandline after expansion, particularly in the case when an
1577 unquoted variable expansion results in an empty value.
1579 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1581 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1582 by renaming away from it. Addresses GH issue 15, reported by
1585 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1587 TL/15 SECURITY: prevent double expansion in math comparison functions
1588 (can expand unsanitized data). Not remotely exploitable.
1595 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1597 PP/02 Make -n do something, by making it not do something.
1598 When combined with -bP, the name of an option is not output.
1600 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1603 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1604 $sender_host_name and config options to manage this, and basic check
1607 PP/05 DSCP support for outbound connections and control modifier for inbound.
1609 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1610 (Only plugin which currently uses this is kerberos4, which nobody should
1611 be using, but we should make it available and other future plugins might
1612 conceivably use it, even though it would break NAT; stuff *should* be
1613 using channel bindings instead).
1615 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1616 name; added for Sendmail compatibility; requires admin caller.
1617 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1618 just ignore it); requires trusted caller.
1619 Also parse but ignore: -Ac -Am -X<logfile>
1622 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1624 TL/02 Add +smtp_confirmation as a default logging option.
1626 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1627 Patch by Magnus Holmgren from 2007-02-20.
1629 TL/04 Bugzilla 1281 - Spec typo.
1630 Bugzilla 1283 - Spec typo.
1631 Bugzilla 1290 - Spec grammar fixes.
1633 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1635 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1637 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1638 mailing list by by Dmitry Isaikin.
1640 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1642 JH/02 Support "G" suffix to numbers in ${if comparisons.
1644 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1646 NM/01 Bugzilla 1197 - Spec typo
1647 Bugzilla 1196 - Spec examples corrections
1649 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1651 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1652 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1655 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1656 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1658 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1659 "acl {{name}{arg}...}", and optional args on acl condition
1662 JH/05 Permit multiple router/transport headers_add/remove lines.
1664 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1666 JH/07 Avoid using a waiting database for a single-message-only transport.
1667 Performance patch from Paul Fisher. Bugzilla 1262.
1669 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1672 JH/09 Add $headers_added variable, with content from use of ACL modifier
1673 add_header (but not yet added to the message). Bugzilla 199.
1675 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1676 Pulled from Bugzilla 817 by Wolfgang Breyha.
1678 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1680 (nb: this is the same fix as in Exim 4.80.1)
1682 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1685 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1688 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1690 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1691 advertise SMTP AUTH mechanism to us, instead of a generic
1692 protocol violation error. Also, make Exim more robust to bad
1693 data from the Dovecot auth socket.
1695 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1697 When a queue runner is handling a message, Exim first routes the
1698 recipient addresses, during which it prunes them based on the retry
1699 hints database. After that it attempts to deliver the message to
1700 any remaining recipients. It then updates the hints database using
1703 So if a recipient address works intermittently, it can get repeatedly
1704 deferred at routing time. The retry hints record remains fresh so the
1705 address never reaches the final cutoff time.
1707 This is a fairly common occurrence when a user is bumping up against
1708 their storage quota. Exim had some logic in its local delivery code
1709 to deal with this. However it did not apply to per-recipient defers
1710 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1712 This change adds a proper retry rule check during routing so that the
1713 final cutoff time is checked against the message's age. We only do
1714 this check if there is an address retry record and there is not a
1715 domain retry record; this implies that previous attempts to handle
1716 the address had the retry_use_local_parts option turned on. We use
1717 this as an approximation for the destination being like a local
1718 delivery, as in LMTP.
1720 I suspect this new check makes the old local delivery cutoff check
1721 redundant, but I have not verified this so I left the code in place.
1723 TF/02 Correct gecos expansion when From: is a prefix of the username.
1725 Test 0254 submits a message to Exim with the header
1729 When I ran the test suite under the user fanf2, Exim expanded
1730 the header to contain my full name, whereas it should have added
1731 a Resent-Sender: header. It erroneously treats any prefix of the
1732 username as equal to the username.
1734 This change corrects that bug.
1736 GF/01 DCC debug and logging tidyup
1737 Error conditions log to paniclog rather than rejectlog.
1738 Debug lines prefixed by "DCC: " to remove any ambiguity.
1740 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1742 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1743 Bug spotted by Jeremy Harris; was flawed since initial commit.
1744 Would have resulted in OCSP responses post-SNI triggering an Exim
1745 NULL dereference and crash.
1747 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1749 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1750 Bug detection, analysis and fix by Samuel Thibault.
1751 Bugzilla 1331, Debian bug #698092.
1753 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1755 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1756 Server implementation by Todd Lyons, client by JH.
1757 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1758 config variable "prdr_enable" controls whether the server
1759 advertises the facility. If the client requests PRDR a new
1760 acl_data_smtp_prdr ACL is called once for each recipient, after
1761 the body content is received and before the acl_smtp_data ACL.
1762 The client is controlled by both of: a hosts_try_prdr option
1763 on the smtp transport, and the server advertisement.
1764 Default client logging of deliveries and rejections involving
1765 PRDR are flagged with the string "PRDR".
1767 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1768 fclose(). Diagnosis by Todd Lyons.
1770 PP/17 Update configure.default to handle IPv6 localhost better.
1771 Patch by Alain Williams (plus minor tweaks).
1774 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1775 This is now consistent with GnuTLS, and is now documented: the
1776 previous undocumented portable approach to treating the option as
1777 unset was to force an expansion failure. That still works, and
1778 an empty string is now equivalent.
1780 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1781 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1782 not performing validation itself.
1784 PP/20 Added force_command boolean option to pipe transport.
1785 Patch from Nick Koston, of cPanel Inc.
1787 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1790 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1792 PP/21 Fix eximon continuous updating with timestamped log-files.
1793 Broken in a format-string cleanup in 4.80, missed when I repaired the
1794 other false fix of the same issue.
1795 Report and fix from Heiko Schlichting.
1798 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1799 Report from Prashanth Katuri.
1801 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1802 It's SecureTransport, so affects any MacOS clients which use the
1803 system-integrated TLS libraries, including email clients.
1805 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1806 using a MIME ACL for non-SMTP local injection.
1807 Report and assistance in diagnosis by Warren Baker.
1809 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1811 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1813 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1814 last id that failed so it may be referenced in subsequent ACL's.
1816 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1819 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1820 ldap library initialization, allowing self-signed CA's to be
1821 used. Also properly sets require_cert option later in code by
1822 using NULL (global ldap config) instead of ldap handle (per
1823 session). Bug diagnosis and testing by alxgomz.
1825 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1826 the src/util/ subdirectory.
1828 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1829 renamed to Transport Post Delivery Action by Jeremy Harris, as
1832 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1833 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1834 redis_servers = needs to be configured which will be used by the redis
1835 lookup. Patch from Warren Baker, of The Packet Hub.
1837 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1839 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1840 hostname or reverse DNS when processing a host list. Used suggestions
1841 from multiple comments on this bug.
1843 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1845 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1846 Missed a few lines, added it to make the runtest require no keyboard
1849 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1850 contains upper case chars. Make router use caseful_local_part.
1852 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1853 support when GnuTLS has been built with p11-kit.
1859 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1861 This, or similar/improved, will also be change PP/11 of 4.82.
1867 PP/01 Handle short writes when writing local log-files.
1868 In practice, only affects FreeBSD (8 onwards).
1869 Bugzilla 1053, with thanks to Dmitry Isaikin.
1871 NM/01 Bugzilla 949 - Documentation tweak
1873 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1876 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1878 PP/02 Implemented gsasl authenticator.
1880 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1882 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1883 `pkg-config foo` for cflags/libs.
1885 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1886 with rest of GSASL and with heimdal_gssapi.
1888 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1889 `pkg-config foo` for cflags/libs for the TLS implementation.
1891 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1892 properties get this fed in as external SSF. A number of robustness
1893 and debugging improvements to the cyrus_sasl authenticator.
1895 PP/08 cyrus_sasl server now expands the server_realm option.
1897 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1898 Patch by Jeremy Harris.
1900 PP/10 Added dbmjz lookup type.
1902 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1904 PP/12 MAIL args handles TAB as well as SP, for better interop with
1905 non-compliant senders.
1906 Analysis and variant patch by Todd Lyons.
1908 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1909 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1910 Patch from Dirk Mueller <dmueller@suse.com>
1912 PP/13 tls_peerdn now print-escaped for spool files.
1913 Observed some $tls_peerdn in wild which contained \n, which resulted
1914 in spool file corruption.
1916 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1917 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1918 or write after TLS renegotiation, which otherwise led to messages
1921 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1922 as a tracking header (ie: a signed header comes before the signature).
1923 Patch from Wolfgang Breyha.
1925 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1926 comma-sep list; embedded commas doubled.
1928 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1930 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1932 Report and patch from Dmitry Banschikov.
1934 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1935 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1936 failures. We appear to now support TLS1.1+ with Exim.
1938 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1939 lets Exim select keys and certificates based upon TLS SNI from client.
1940 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1941 before an outbound SMTP session. New log_selector, +tls_sni.
1943 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1944 NULL dereference. Report and patch from Alun Jones.
1946 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1947 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1948 Not seeing resolver debug output on NetBSD, but suspect this is a
1949 resolver implementation change.
1951 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1952 Left warnings. Added "eximon gdb" invocation mode.
1954 PP/21 Defaulting "accept_8bitmime" to true, not false.
1956 PP/22 Added -bw for inetd wait mode support.
1958 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1959 locate the relevant includes and libraries. Made this the default.
1961 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1962 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1964 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1965 This may cause build issues on older platforms.
1967 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1968 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1969 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1970 Added SNI support via GnuTLS too.
1971 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1973 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1975 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1976 Applied second patch from Janne, implementing suggestion to default
1977 multiple-strings-in-record handling to match SPF spec.
1979 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1981 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1982 read-only, out of scope).
1983 Patch from Wolfgang Breyha, report from Stuart Northfield.
1985 PP/29 Fix three issues highlighted by clang analyser static analysis.
1986 Only crash-plausible issue would require the Cambridge-specific
1987 iplookup router and a misconfiguration.
1988 Report from Marcin Mirosław.
1990 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1992 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1993 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1994 As part of this, removing so much warning spew let me fix some minor
1995 real issues in debug logging.
1997 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1998 assignment on my part. Fixed.
2000 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2001 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2002 Janne Snabb (who went above and beyond: thank you).
2004 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2005 string otherwise requires a connection and a bunch more work and it's
2006 relatively easy to get wrong. Should also expose TLS library linkage
2009 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2010 64-bit ${eval} (JH/03).
2012 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2013 GNU libc to support some of the 64-bit stuff, should not lead to
2014 conflicts. Defined before os.h is pulled in, so if a given platform
2015 needs to override this, it can.
2017 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2018 protection layer was required, which is not implemented.
2019 Bugzilla 1254, patch from Wolfgang Breyha.
2021 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2022 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2023 tls_dhparam take prime identifiers. Also unbreak combination of
2024 OpenSSL+DH_params+TLSSNI.
2026 PP/39 Disable SSLv2 by default in OpenSSL support.
2032 PP/01 Solaris build fix for Oracle's LDAP libraries.
2033 Bugzilla 1109, patch from Stephen Usher.
2035 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2037 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2040 TF/02 Fix a couple more cases where we did not log the error message
2041 when unlink() failed. See also change 4.74-TF/03.
2043 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2044 lock up or crash if it happened to be inside a call to libc when it
2045 got a SIGUSR1 from exiwhat.
2047 The SIGUSR1 handler appends the current process status to the process
2048 log which is later printed by exiwhat. It used to use the general
2049 purpose logging code to do this, but several functions it calls are
2050 not safe for signals.
2052 The new output code in the SIGUSR1 handler is specific to the process
2053 log, and simple enough that it's easy to inspect for signal safety.
2054 Removing some special cases also simplifies the general logging code.
2055 Removing the spurious timestamps from the process log simplifies
2058 TF/04 Improved ratelimit ACL condition.
2060 The /noupdate option has been deprecated in favour of /readonly which
2061 has clearer semantics. The /leaky, /strict, and /readonly update modes
2062 are mutually exclusive. The update mode is no longer included in the
2063 database key; it just determines when the database is updated. (This
2064 means that when you upgrade Exim will forget old rate measurements.)
2066 Exim now checks that the per_* options are used with an update mode that
2067 makes sense for the current ACL. For example, when Exim is processing a
2068 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2069 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2070 must specify per_mail/readonly. If you omit the update mode it defaults to
2071 /leaky where that makes sense (as before) or /readonly where required.
2073 The /noupdate option is now undocumented but still supported for
2074 backwards compatibility. It is equivalent to /readonly except that in
2075 ACLs where /readonly is required you may specify /leaky/noupdate or
2076 /strict/noupdate which are treated the same as /readonly.
2078 A useful new feature is the /count= option. This is a generalization
2079 of the per_byte option, so that you can measure the throughput of other
2080 aggregate values. For example, the per_byte option is now equivalent
2081 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2083 The per_rcpt option has been generalized using the /count= mechanism
2084 (though it's more complicated than the per_byte equivalence). When it is
2085 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2086 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2087 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2088 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2089 Note that using per_rcpt with a non-readonly update mode in more than
2090 one ACL will cause the recipients to be double-counted. (The per_mail
2091 and per_byte options don't have this problem.)
2093 The handling of very low rates has changed slightly. If the computed rate
2094 is less than the event's count (usually one) then this event is the first
2095 after a long gap. In this case the rate is set to the same as this event's
2096 count, so that the first message of a spam run is counted properly.
2098 The major new feature is a mechanism for counting the rate of unique
2099 events. The new per_addr option counts the number of different
2100 recipients that someone has sent messages to in the last time period. It
2101 behaves like per_rcpt if all the recipient addresses are different, but
2102 duplicate recipient addresses do not increase the measured rate. Like
2103 the /count= option this is a general mechanism, so the per_addr option
2104 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2105 example, measure the rate that a client uses different sender addresses
2106 with the options per_mail/unique=$sender_address. There are further
2107 details in the main documentation.
2109 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2111 TF/06 Removed a few PCRE remnants.
2113 TF/07 Automatically extract Exim's version number from tags in the git
2114 repository when doing development or release builds.
2116 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2117 Bugzilla 879. Patch from Paul Fisher.
2119 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2120 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2123 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2125 PP/05 Variable $av_failed, true if the AV scanner deferred.
2126 Bugzilla 1078. Patch from John Horne.
2128 PP/06 Stop make process more reliably on build failure.
2129 Bugzilla 1087. Patch from Heiko Schlittermann.
2131 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2132 Bugzilla 1089. Patch from Heiko Schlittermann.
2134 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2135 Bugzilla 1131. Patch from Holger Weiß.
2137 PP/09 Handle IPv6 addresses with SPF.
2138 Bugzilla 860. Patch from Wolfgang Breyha.
2140 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2142 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2145 PP/11 match_* no longer expand right-hand-side by default.
2146 New compile-time build option, EXPAND_LISTMATCH_RHS.
2147 New expansion conditions, "inlist", "inlisti".
2149 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2151 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2153 PP/14 fix log_write() format string regression from TF/03.
2154 Bugzilla 1152. Patch from Dmitry Isaikin.
2160 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2162 PP/02 Harmonised TLS library version reporting; only show if debugging.
2163 Layout now matches that introduced for other libraries in 4.74 PP/03.
2165 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2167 PP/04 New "dns_use_edns0" global option.
2169 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2172 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2173 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2175 TK/01 Updated PolarSSL code to 0.14.2.
2176 Bugzilla 1097. Patch from Andreas Metzler.
2178 PP/07 Catch divide-by-zero in ${eval:...}.
2179 Fixes bugzilla 1102.
2181 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2184 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2185 format-string attack -- SECURITY: remote arbitrary code execution.
2187 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2188 time unintentionally subject to list matching rules, letting the header
2189 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2190 arbitrary string expansion). This allowed for information disclosure.
2192 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2193 INT_MIN/-1 -- value coerced to INT_MAX.
2199 NM/01 Workaround for PCRE version dependency in version reporting
2202 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2203 This fixes portability to compilers other than gcc, notably
2204 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2206 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2207 makefiles for portability to HP-UX and POSIX correctness.
2209 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2210 Also via indented variable definition in the Makefile.
2211 (Debugging by Oliver Heesakkers).
2213 PP/02 Restore caching of spamd results with expanded spamd_address.
2214 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2216 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2217 Improves build reliability. Fix from: Frank Elsner
2219 NM/02 Fix wide character breakage in the rfc2047 coding
2220 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2222 NM/03 Allow underscore in dnslist lookups
2223 Fixes bug 1026. Patch from Graeme Fowler
2225 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2226 Code patches from Adam Ciarcinski of NetBSD.
2228 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2231 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2232 is logged, to avoid truncation. Patch from John Horne.
2234 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2235 Patch from Jakob Hirsch.
2237 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2238 SQL string expansion failure details.
2239 Patch from Andrey Oktyabrski.
2241 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2242 Patch from Simon Arlott.
2244 PP/09 New lookups functionality failed to compile on old gcc which rejects
2245 extern declarations in function scope.
2246 Patch from Oliver Fleischmann
2248 PP/10 Use sig_atomic_t for flags set from signal handlers.
2249 Check getgroups() return and improve debugging.
2250 Fixed developed for diagnosis in bug 927 (which turned out to be
2253 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2254 Patch from Mark Zealey.
2256 PP/12 Bugzilla 1056: Improved spamd server selection.
2257 Patch from Mark Zealey.
2259 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2260 Based on patch from Heiko Schlittermann.
2262 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2263 Patch from Uwe Doering, sign-off by Michael Haardt.
2265 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2266 variable declaration deep within a block. Bug and patch from
2269 PP/15 lookups-Makefile IRIX compatibility coercion.
2271 PP/16 Make DISABLE_DKIM build knob functional.
2273 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2274 Patch by Simon Arlott
2276 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2277 variable argument macros. Our copy now differs from upstream.
2283 TF/01 Failure to get a lock on a hints database can have serious
2284 consequences so log it to the panic log.
2286 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2287 controlled using the smtp_confirmation log selector.
2289 TF/03 Include the error message when we fail to unlink a spool file.
2291 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2292 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2293 for maintaining out-of-tree patches for some time.
2295 PP/01 Bugzilla 139: Documentation and portability issues.
2296 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2297 Handle per-OS dynamic-module compilation flags.
2299 PP/02 Let /dev/null have normal permissions.
2300 The 4.73 fixes were a little too stringent and complained about the
2301 permissions on /dev/null. Exempt it from some checks.
2302 Reported by Andreas M. Kirchwitz.
2304 PP/03 Report version information for many libraries, including
2305 Exim version information for dynamically loaded libraries. Created
2306 version.h, now support a version extension string for distributors
2307 who patch heavily. Dynamic module ABI change.
2309 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2310 privilege escalation vulnerability whereby the Exim run-time user
2311 can cause root to append content of the attacker's choosing to
2314 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2317 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2318 If dropping privileges for untrusted macros, we disabled normal logging
2319 on the basis that it would fail; for the Exim run-time user, this is not
2320 the case, and it resulted in successful deliveries going unlogged.
2321 Fixed. Reported by Andreas Metzler.
2327 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2328 only prepend for the Resent-* case. Fixes regression introduced in
2329 Exim 4.70 by NM/22 for Bugzilla 607.
2331 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2332 increasing numbers of administrators be bitten by this.
2334 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2336 PP/03 Bugzilla 994: added openssl_options main configuration option.
2338 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2340 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2342 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2344 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2345 without a peer certificate, leading to a segfault because of an
2346 assumption that peers always have certificates. Be a little more
2347 paranoid. Problem reported by Martin Tscholak.
2349 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2350 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2351 NB: ClamAV planning to remove STREAM in "middle of 2010".
2352 CL also introduces -bmalware, various -d+acl logging additions and
2353 more caution in buffer sizes.
2355 PP/09 Implemented reverse_ip expansion operator.
2357 PP/10 Bugzilla 937: provide a "debug" ACL control.
2359 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2361 PP/12 Bugzilla 973: Implement --version.
2363 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2365 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2367 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2369 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2370 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2371 ignore trailing whitespace.
2373 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2375 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2378 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2379 Notification from Dr Andrew Aitchison.
2381 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2382 ExtendedDetectionInfo response format.
2383 Notification from John Horne.
2385 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2388 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2389 XSL and documented dependency on system catalogs, with examples of how
2392 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2395 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2396 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2397 configuration file which is writeable by the Exim user or group.
2399 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2400 of configuration files to cover files specified with the -C option if
2401 they are going to be used with root privileges, not just the default
2404 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2405 option (effectively making it always true).
2407 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2408 files to be used while preserving root privileges.
2410 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2411 that rogue child processes cannot use them.
2413 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2414 run-time user, instead of root.
2416 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2417 Exim run-time user without dropping privileges.
2419 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2420 result string, instead of calling string_vformat() twice with the same
2423 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2424 for other users. Others should always drop root privileges if they use
2425 -C on the command line, even for a whitelisted configure file.
2427 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2429 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2435 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2436 $data_path, and $header_path variables; fixed documentation bugs and
2439 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2440 exipick to access non-standard spools, including the "frozen" queue
2443 NM/01 Bugzilla 965: Support mysql stored procedures.
2444 Patch from Alain Williams
2446 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2448 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2449 Patch from Andreas Metzler
2451 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2452 Patch from Kirill Miazine
2454 NM/05 Bugzilla 671: Added umask to procmail example.
2456 JJ/03 installed exipick 20100323.0, fixing doc bug
2458 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2459 directory. Notification and patch from Dan Rosenberg.
2461 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2463 TK/02 Improve log output when DKIM signing operation fails.
2465 MH/01 Treat the transport option dkim_domain as a colon separated
2466 list, not as a single string, and sign the message with each element,
2467 omitting multiple occurences of the same signer.
2469 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2470 Bugzilla 985, 986. Patch by Simon Arlott
2472 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2473 Patch by Simon Arlott
2475 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2476 MBX locking. Notification from Dan Rosenberg.
2482 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2484 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2486 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2488 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2490 NM/04 Bugzilla 915: Flag broken perl installation during build.
2496 TK/01 Added patch by Johannes Berg that expands the main option
2497 "spamd_address" if it starts with a dollar sign.
2499 TK/02 Write list of recipients to X-Envelope-Sender header when building
2500 the mbox-format spool file for content scanning (suggested by Jakob
2503 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2504 (http://www.dcc-servers.net/) support via dccifd. Activated by
2505 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2507 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2508 by Mark Daniel Reidel <mr@df.eu>.
2510 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2511 When building exim an external PCRE library is now needed -
2512 PCRE is a system library on the majority of modern systems.
2513 See entry on PCRE_LIBS in EDITME file.
2515 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2516 conversation. Added nologin parameter to request.
2517 Patch contributed by Kirill Miazine.
2519 TF/01 Do not log submission mode rewrites if they do not change the address.
2521 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2523 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2524 log files in place. Contributed by Roberto Lima.
2526 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2528 TF/03 Bugzilla 615: When checking the local_parts router precondition
2529 after a local_part_suffix or local_part_prefix option, Exim now
2530 does not use the address's named list lookup cache, since this
2531 contains cached lookups for the whole local part.
2533 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2534 Robert Millan. Documentation is in experimental-spec.txt.
2536 TF/04 Bugzilla 668: Fix parallel build (make -j).
2538 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2540 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2541 Patch provided by Jan Srzednicki.
2543 TF/05 Leading white space used to be stripped from $spam_report which
2544 wrecked the formatting. Now it is preserved.
2546 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2547 that they are available at delivery time.
2549 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2551 TF/08 TLS error reporting now respects the incoming_interface and
2552 incoming_port log selectors.
2554 TF/09 Produce a more useful error message if an SMTP transport's hosts
2555 setting expands to an empty string.
2557 NM/06 Bugzilla 744: EXPN did not work under TLS.
2558 Patch provided by Phil Pennock.
2560 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2561 Patch provided by Richard Godbee.
2563 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2564 acl_smtp_notquit, added index entry.
2566 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2567 Patch provided by Eugene Bujak.
2569 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2570 accept(). Patch provided by Maxim Dounin.
2572 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2573 Patch provided by Phil Pennock.
2575 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2577 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2578 Patch provided by Brad "anomie" Jorsch.
2580 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2581 Patch provided by Dean Brooks.
2583 TK/05 Add native DKIM support (does not depend on external libraries).
2585 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2586 Patch provided by Graeme Fowler.
2588 NM/16 Bugzilla 851: Documentation example syntax fix.
2590 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2592 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2595 NM/19 Bugzilla 745: TLS version reporting.
2596 Patch provided by Phil Pennock.
2598 NM/20 Bugzilla 167: bool: condition support.
2599 Patch provided by Phil Pennock.
2601 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2602 clients. Patch provided by Phil Pennock.
2604 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2605 Patch provided by Brad "anomie" Jorsch.
2607 NM/23 Bugzilla 687: Fix misparses in eximstats.
2608 Patch provided by Heiko Schlittermann.
2610 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2611 Patch provided by Heiko Schlittermann.
2613 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2614 plus update to original patch.
2616 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2618 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2619 Patch provided by David Brownlee.
2621 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2623 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2625 NM/30 Bugzilla 888: TLS documentation bugfixes.
2627 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2629 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2630 Unlike the original bugzilla I have changed all shell scripts in src tree.
2632 NM/33 Bugzilla 898: Transport filter timeout fix.
2633 Patch by Todd Rinaldo.
2635 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2636 Patch by Serge Demonchaux.
2638 NM/35 Bugzilla 39: Base64 decode bug fixes.
2639 Patch by Jakob Hirsch.
2641 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2643 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2645 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2647 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2653 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2654 ALT-N's libdkim that I have put here:
2655 http://duncanthrax.net/exim-experimental/
2657 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2658 were called 'true' and it seems that C99 defines that as a reserved
2659 keyword to be used with 'bool' variable types. That means you could
2660 not include C99-style headers which use bools without triggering
2661 build errors in sieve.c.
2663 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2664 as mailq or other aliases. Changed the --help handling significantly
2665 to do whats expected. exim_usage() emits usage/help information.
2667 SC/01 Added the -bylocaldomain option to eximstats.
2669 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2671 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2673 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2679 PH/01 Another patch from the Sieve maintainer.
2681 PH/02 When an IPv6 address is converted to a string for single-key lookup
2682 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2683 dots are used instead of colons so that keys in lsearch files need not
2684 contain colons. This was done some time before quoting was made available
2685 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2686 (notated using the quote facility) so as to distinguish them from IPv4
2687 keys. This meant that lookups for IP addresses in host lists did not work
2688 for iplsearch lookups.
2690 This has been fixed by arranging for IPv6 addresses to be expressed with
2691 colons if the lookup type is iplsearch. This is not incompatible, because
2692 previously such lookups could never work.
2694 The situation is now rather anomalous, since one *can* have colons in
2695 ordinary lsearch keys. However, making the change in all cases is
2696 incompatible and would probably break a number of configurations.
2698 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2701 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2702 conversion specification without a maximum field width, thereby enabling
2703 a rogue spamd server to cause a buffer overflow. While nobody in their
2704 right mind would setup Exim to query an untrusted spamd server, an
2705 attacker that gains access to a server running spamd could potentially
2706 exploit this vulnerability to run arbitrary code as the Exim user.
2708 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2709 $primary_hostname instead of what libspf2 thinks the hosts name is.
2711 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2712 a directory entry by the name of the lookup key. Previously, if a
2713 symlink pointed to a non-existing file or a file in a directory that
2714 Exim lacked permissions to read, a lookup for a key matching that
2715 symlink would fail. Now it is enough that a matching directory entry
2716 exists, symlink or not. (Bugzilla 503.)
2718 PH/03 The body_linecount and body_zerocount variables are now exported in the
2721 PH/04 Added the $dnslist_matched variable.
2723 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2724 This means they are set thereafter only if the connection becomes
2727 PH/06 Added the client_condition to authenticators so that some can be skipped
2728 by clients under certain conditions.
2730 PH/07 The error message for a badly-placed control=no_multiline_responses left
2731 "_responses" off the end of the name.
2733 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2735 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2736 (without spaces) instead of just copying the configuration text.
2738 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2740 PH/11 Added $max_received_linelength.
2742 PH/12 Added +ignore_defer and +include_defer to host lists.
2744 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2745 way in which PCRE > 7.0 is built.
2747 PH/14 Implemented queue_only_load_latch.
2749 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2750 MAIL command. The effect was to mangle the value on 64-bit systems.
2752 PH/16 Another patch from the Sieve maintainer.
2754 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2756 PH/18 If a system quota error occurred while trying to create the file for
2757 a maildir delivery, the message "Mailbox is full" was not appended to the
2758 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2759 only to a quota excession during the actual writing of the file.
2761 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2762 characters?) which causes problems in log lines. The DN values are now
2763 passed through string_printing() before being added to log lines.
2765 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2766 and InterBase are left for another time.)
2768 PH/21 Added message_body_newlines option.
2770 PH/22 Guard against possible overflow in moan_check_errorcopy().
2772 PH/23 POSIX allows open() to be a macro; guard against that.
2774 PH/24 If the recipient of an error message contained an @ in the local part
2775 (suitably quoted, of course), incorrect values were put in $domain and
2776 $local_part during the evaluation of errors_copy.
2782 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2783 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2786 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2787 issue a MAIL command.
2789 PH/02 In an ACL statement such as
2791 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2793 if a client was not listed at all, or was listed with a value other than
2794 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2795 the condition was not true (as it should be), so access was not denied.
2796 The bug was that the ! inversion was incorrectly passed on to the second
2797 item. This has been fixed.
2799 PH/03 Added additional dnslists conditions == and =& which are different from
2800 = and & when the dns lookup returns more than one IP address.
2802 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2803 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2805 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2806 FSYNC, which compiles an option called disable_fsync that allows for
2807 bypassing fsync(). The documentation is heavily laced with warnings.
2809 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2811 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2812 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2813 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2814 including adding "make clean"; (3) Added -fPIC when compiling the test
2815 dynamically loaded module, to get rid of a warning.
2817 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2818 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2819 = 0s. The bug is otherwise harmless.
2821 PH/07 There was a bug in the dovecot authenticator such that the value of
2822 $auth1 could be overwritten, and so not correctly preserved, after a
2823 successful authentication. This usually meant that the value preserved by
2824 the server_setid option was incorrect.
2826 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2828 PH/09 Installed PCRE release 7.0.
2830 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2831 run for batched SMTP input. It is now run at the start of every message
2832 in the batch. While fixing this I discovered that the process information
2833 (output by running exiwhat) was not always getting set for -bs and -bS
2834 input. This is fixed, and it now also says "batched" for BSMTP.
2836 PH/11 Added control=no_pipelining.
2838 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2839 patch, slightly modified), and move the expansion of helo_data till after
2840 the connection is made in the smtp transport (so it can use these
2843 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2845 PH/14 Added log_selector = +pid.
2847 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2849 PH/16 Add ${if forany and ${if forall.
2851 PH/17 Added dsn_from option to vary the From: line in DSNs.
2853 PH/18 Flush SMTP output before performing a callout, unless control =
2854 no_callout_flush is set.
2856 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2857 was true (the default) a successful delivery failed to delete the retry
2858 item, thus causing premature timeout of the address. The bug is now
2861 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2863 PH/21 Long custom messages for fakedefer and fakereject are now split up
2864 into multiline responses in the same way that messages for "deny" and
2865 other ACL rejections are.
2867 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2868 with slight modification.
2870 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2871 draft, changing the syntax and factoring some duplicate code".
2873 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2874 for deliveries of the second and subsequent messages over the same SMTP
2877 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2878 ${reduce, with only minor "tidies".
2880 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2882 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2883 expansion side effects.
2885 PH/27 When a message times out after an over-quota error from an Exim-imposed
2886 quota, the bounce message says "mailbox is full". This message was not
2887 being given when it was a system quota that was exceeded. It now should
2890 MH/03 Made $recipients available in local_scan(). local_scan() already has
2891 better access to the recipient list through recipients_list[], but
2892 $recipients can be useful in postmaster-provided expansion strings.
2894 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2895 in the case of a MAIL command with additional options following the
2896 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2897 were accidentally chopped off.
2899 PH/29 SMTP synchronization checks are implemented when a command is read -
2900 there is a check that no more input is waiting when there shouldn't be
2901 any. However, for some commands, a delay in an ACL can mean that it is
2902 some time before the response is written. In this time, more input might
2903 arrive, invalidly. So now there are extra checks after an ACL has run for
2904 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2905 pipelining has not been advertised.
2907 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2909 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2910 returned in the SMTP response when smtp_return_error_details was false.
2911 This has been fixed.
2913 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2914 instead of the C I/O that was originally supplied, because problems were
2915 reported on Solaris.
2917 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2918 Exim which did not show up earlier: it was assuming that a call to
2919 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2920 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2921 was a macro that became an assignment, so it seemed to work. This has
2922 changed to a proper function call with a void return, hence the compile
2923 error. Exim's code has been fixed.
2925 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2928 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2930 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2932 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2933 because it turns out to be too big or there is a timeout) there is no
2934 "Completed" line in the log. When some messages of this type were
2935 selected by exigrep, they were listed as "not completed". Others were
2936 picked up by some special patterns. I have improved the selection
2937 criteria to be more general.
2939 PH/38 The host_find_failed option in the manualroute router can now be set
2940 to "ignore", to completely ignore a host whose IP address cannot be
2941 found. If all hosts are ignored, the behaviour is controlled by the new
2942 host_all_ignored option.
2944 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2945 homing or because of multiple MX records with /mx) generated more than
2946 one IP address, and the following item turned out to be the local host,
2947 all the secondary addresses of the first item were incorrectly removed
2948 from the list, along with the local host and any following hosts (which
2949 is what is supposed to happen).
2951 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2952 whoever called Exim into the -H file. In the case of the daemon it was
2953 behaving confusingly. When first started, it used values for whoever
2954 started the daemon, but after a SIGHUP it used the Exim user (because it
2955 calls itself on a restart). I have changed the code so that it now always
2958 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2959 message are rejected with the same error (e.g. no authentication or bad
2960 sender address), and a DATA command is nevertheless sent (as can happen
2961 with PIPELINING or a stupid MUA), the error message that was given to the
2962 RCPT commands is included in the rejection of the DATA command. This is
2963 intended to be helpful for MUAs that show only the final error to their
2966 PH/42 Another patch from the Sieve maintainer.
2968 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2969 Eximstats - Fixed some broken HTML links and added missing column headers
2971 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2972 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2974 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2976 PH/43 Yet another patch from the Sieve maintainer.
2978 PH/44 I found a way to check for a TCP/IP connection going away before sending
2979 the response to the final '.' that terminates a message, but only in the
2980 case where the client has not sent further data following the '.'
2981 (unfortunately, this is allowed). However, in many cases there won't be
2982 any further data because there won't be any more messages to send. A call
2983 to select() can be used: if it shows that the input is "ready", there is
2984 either input waiting, or the socket has been closed. An attempt to read
2985 the next input character can distinguish the two cases. Previously, Exim
2986 would have sent an OK response which the client would never have see.
2987 This could lead to message repetition. This fix should cure that, at
2988 least in a lot of common cases.
2990 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2991 advertised in response to EHLO.
2997 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2998 fixed by 4.65/MH/01 (is this a record?) are fixed:
3000 (i) An empty string was always treated as zero by the numeric comparison
3001 operators. This behaviour has been restored.
3003 (ii) It is documented that the numeric comparison operators always treat
3004 their arguments as decimal numbers. This was broken in that numbers
3005 starting with 0 were being interpreted as octal.
3007 While fixing these problems I realized that there was another issue that
3008 hadn't been noticed. Values of message_size_limit (both the global option
3009 and the transport option) were treated as octal if they started with 0.
3010 The documentation was vague. These values are now always treated as
3011 decimal, and I will make that clear in the documentation.
3017 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3018 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3021 MH/01 Don't check that the operands of numeric comparison operators are
3022 integers when their expansion is in "skipping" mode (fixes bug
3023 introduced by 4.64-PH/07).
3025 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3026 child addresses, Exim now panics and dies. Previously, because the count
3027 is held in a short int, deliveries were likely to be lost. As such a
3028 large number of recipients for a single message is ridiculous
3029 (performance will be very, very poor), I have chosen to impose a limit
3030 rather than extend the field.
3036 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3037 leftover -K file (the existence of which was triggered by #402).
3038 While we were at it, introduced process PID as part of the -K
3039 filename. This should rule out race conditions when creating
3042 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3043 processing considerably. Previous code took too long for large mails,
3044 triggering a timeout which in turn triggers #401.
3046 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3047 in the DK code in transports.c. sendfile() is not really portable,
3048 hence the _LINUX specificness.
3050 TF/01 In the add_headers option to the mail command in an Exim filter,
3051 there was a bug that Exim would claim a syntax error in any
3052 header after the first one which had an odd number of characters
3055 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3056 callout verification, Exim cached a "reject" for the entire domain. This
3057 is correct for most verifications, but it is not correct for a recipient
3058 verification with use_sender or use_postmaster set, because in that case
3059 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3060 case of MAIL FROM:<> rejection from other early rejections (e.g.
3061 rejection of HELO). When verifying a recipient using a non-null MAIL
3062 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3063 Whatever the result of the callout, the value of the domain cache is
3064 left unchanged (for any other kind of callout, getting as far as trying
3065 RCPT means that the domain itself is ok).
3067 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3070 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3071 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3072 session, the daemon ignores ECONNECT errors and logs others; it now
3073 ignores EPIPE as well.
3075 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3076 (quoted-printable decoding).
3078 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3079 later the small subsequent patch to fix an introduced bug.
3081 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3083 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3085 PH/08 An error is now given if message_size_limit is specified negative.
3087 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3088 to be given (somewhat) arbitrary names.
3090 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3093 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3094 miscellaneous code fixes
3096 PH/10 Added the log_reject_target ACL modifier to specify where to log
3099 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3100 hostname. This is wrong, because it relates to the incoming message (and
3101 probably the interface on which it is arriving) and not to the outgoing
3102 callout (which could be using a different interface). This has been
3103 changed to use the value of the helo_data option from the smtp transport
3104 instead - this is what is used when a message is actually being sent. If
3105 there is no remote transport (possible with a router that sets up host
3106 addresses), $smtp_active_hostname is used.
3108 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3109 tweaks were necessary in order to get it to work (see also 21 below):
3110 (a) The code assumed that strncpy() returns a negative number on buffer
3111 overflow, which isn't the case. Replaced with Exim's string_format()
3113 (b) There were several signed/unsigned issues. I just did the minimum
3114 hacking in of casts. There is scope for a larger refactoring.
3115 (c) The code used strcasecmp() which is not a standard C function.
3116 Replaced with Exim's strcmpic() function.
3117 (d) The code set only $1; it now sets $auth1 as well.
3118 (e) A simple test gave the error "authentication client didn't specify
3119 service in request". It would seem that Dovecot has changed its
3120 interface. Fortunately there's a specification; I followed it and
3121 changed what the client sends and it appears to be working now.
3123 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3126 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3127 address A is aliased to B and C, where B exists and C does not. Without
3128 -v the output is "A verified" because verification stops after a
3129 successful redirection if more than one address is generated. However,
3130 with -v the child addresses are also verified. Exim was outputting "A
3131 failed to verify" and then showing the successful verification for C,
3132 with its parentage. It now outputs "B failed to verify", showing B's
3133 parentage before showing the successful verification of C.
3135 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3136 look up a TXT record in a specific list after matching in a combined
3139 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3140 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3141 they consult the DNS. I had assumed they would set it the way they
3142 wanted; and indeed my experiments on Linux seem to show that in some
3143 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3144 To be on the safe side, however, I have now made the interface to
3145 host_find_byname() similar to host_find_bydns(), with an argument
3146 containing the DNS resolver options. The host_find_byname() function now
3147 sets these options at its start, just as host_find_bydns() does. The smtp
3148 transport options dns_qualify_single and dns_search_parents are passed to
3149 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3150 of host_find_byname() use the default settings of RES_DEFNAMES
3151 (qualify_single) but not RES_DNSRCH (search_parents).
3153 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3154 spool_read_header() do less string testing, by means of a preliminary
3155 switch on the second character of optional "-foo" lines. (This is
3156 overdue, caused by the large number of possibilities that now exist.
3157 Originally there were few.) While I was there, I also converted the
3158 str(n)cmp tests so they don't re-test the leading "-" and the first
3159 character, in the hope this might squeeze out yet more improvement.
3161 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3162 flag allowing group syntax was set by the header_syntax check but not
3163 turned off, possible causing trouble later; (2) The flag was not being
3164 set at all for the header_verify test, causing "group"-style headers to
3165 be rejected. I have now set it in this case, and also caused header_
3166 verify to ignore an empty address taken from a group. While doing this, I
3167 came across some other cases where the code for allowing group syntax
3168 while scanning a header line wasn't quite right (mostly, not resetting
3169 the flag correctly in the right place). These bugs could have caused
3170 trouble for malformed header lines. I hope it is now all correct.
3172 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3173 with the "reply" argument non-NULL. The code, however (which originally
3174 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3175 but it didn't always do it. This confused somebody who was copying the
3176 code for some other use. I have removed all the tests.
3178 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3179 feature that was used to support insecure browsers during the U.S. crypto
3180 embargo. It requires special client support, and Exim is probably the
3181 only MTA that supported it -- and would never use it because real RSA is
3182 always available. This code has been removed, because it had the bad
3183 effect of slowing Exim down by computing (never used) parameters for the
3184 RSA_EXPORT functionality.
3186 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3187 authenticator to fail if there's a tab character in the incoming data
3188 (there should never be unless someone is messing about, as it's supposed
3189 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3190 if the connection is using TLS or if the remote IP is the same as the
3191 local IP, and the "valid-client-cert option" if a client certificate has
3194 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3195 authenticators. This can be used for authorization after authentication
3196 succeeds. (In the case of plaintext, it servers for both authentication
3199 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3200 if any retry times were supplied.
3202 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3203 connection, where there is no client IP address to check. In this
3204 situation, the verify now always succeeds.
3206 PH/25 Applied John Jetmore's -Mset patch.
3208 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3210 PH/27 In a string expansion for a processed (not raw) header when multiple
3211 headers of the same name were present, leading whitespace was being
3212 removed from all of them, but trailing whitespace was being removed only
3213 from the last one. Now trailing whitespace is removed from each header
3214 before concatenation. Completely empty headers in a concatenation (as
3215 before) are ignored.
3217 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3218 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3220 PH/29 [Removed. This was a change that I later backed out, and forgot to
3221 correct the ChangeLog entry (that I had efficiently created) before
3222 committing the later change.]
3224 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3225 address errors (4xx response to RCPT) over the same connection as other
3226 messages routed to the same hosts. Such deliveries are always "forced",
3227 so retry times are not inspected. This resulted in far too many retries
3228 for the affected addresses. The effect occurred only when there were more
3229 hosts than the hosts_max_try setting in the smtp transport when it had
3230 the 4xx errors. Those hosts that it had tried were not added to the list
3231 of hosts for which the message was waiting, so if all were tried, there
3232 was no problem. Two fixes have been applied:
3234 (i) If there are any address or message errors in an SMTP delivery, none
3235 of the hosts (tried or untried) are now added to the list of hosts
3236 for which the message is waiting, so the message should not be a
3237 candidate for sending over the same connection that was used for a
3238 successful delivery of some other message. This seems entirely
3239 reasonable: after all the message is NOT "waiting for some host".
3240 This is so "obvious" that I'm not sure why it wasn't done
3241 previously. Hope I haven't missed anything, but it can't do any
3242 harm, as the worst effect is to miss an optimization.
3244 (ii) If, despite (i), such a delivery is accidentally attempted, the
3245 routing retry time is respected, so at least it doesn't keep
3246 hammering the server.
3248 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3249 in ${readsocket because some servers need this prod.
3251 PH/32 Added some extra debug output when updating a wait-xxx database.
3253 PH/33 The hint "could be header name not terminated by colon", which has been
3254 given for certain expansion errors for a long time, was not being given
3255 for the ${if def:h_colon_omitted{... case.
3257 PH/34 The spec says: "With one important exception, whenever a domain list is
3258 being scanned, $domain contains the subject domain." There was at least
3259 one case where this was not true.
3261 PH/35 The error "getsockname() failed: connection reset by peer" was being
3262 written to the panic log as well as the main log, but it isn't really
3263 panic-worthy as it just means the connection died rather early on. I have
3264 removed the panic log writing for the ECONNRESET error when getsockname()
3267 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3268 runs only) independently of the message's sender address. This meant
3269 that, if the 4xx error was in fact related to the sender, a different
3270 message to the same recipient with a different sender could confuse
3271 things. In particular, this can happen when sending to a greylisting
3272 server, but other circumstances could also provoke similar problems.
3273 I have changed the default so that the retry time for these errors is now
3274 based a combination of the sender and recipient addresses. This change
3275 can be overridden by setting address_retry_include_sender=false in the
3278 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3279 remote server are returned as part of bounce messages. This was not
3280 happening for LMTP over a pipe (the lmtp transport), but now it is the
3281 same for both kinds of LMTP.
3283 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3284 in header lines that were in fact CNAMEs. This is no longer the case.
3286 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3287 and queue runs started by the daemon processed all messages. This has
3288 been fixed so that -R and -S can now usefully be given with -q<time>.
3290 PH/40 Import PCRE release 6.7 (fixes some bugs).
3292 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3294 PH/42 Give an error if -q is specified more than once.
3296 PH/43 Renamed the variables $interface_address and $interface_port as
3297 $received_ip_address and $received_port, to make it clear that these
3298 values apply to message reception, and not to the outgoing interface when
3299 a message is delivered. (The old names remain recognized, of course.)
3301 PH/44 There was no timeout on the connect() call when using a Unix domain
3302 socket in the ${readsocket expansion. There now is.
3304 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3305 be meaningful with "accept".
3307 SC/01 Eximstats V1.43
3308 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3310 SC/02 Eximstats V1.44
3311 Use a glob alias rather than an array ref in the generated
3312 parser. This improves both readability and performance.
3314 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3315 Collect SpamAssassin and rejection statistics.
3316 Don't display local sender or destination tables unless
3317 there is data to show.
3318 Added average volumes into the top table text output.
3320 SC/04 Eximstats V1.46
3321 Collect data on the number of addresses (recipients)
3322 as well as the number of messages.
3324 SC/05 Eximstats V1.47
3325 Added 'Message too big' to the list of mail rejection
3326 reasons (thanks to Marco Gaiarin).
3328 SC/06 Eximstats V1.48
3329 Mainlog lines which have GMT offsets and are too short to
3330 have a flag are now skipped.
3332 SC/07 Eximstats V1.49 (Alain Williams)
3333 Added the -emptyok flag.
3335 SC/08 Eximstats V1.50
3336 Fixes for obtaining the IP address from reject messages.
3338 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3339 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3340 whitespace changes from 4.64-PH/27
3342 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3345 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3346 are found, allow negative numbers in numeric criteria)
3348 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3350 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3351 to match changes made in 4.64-PH/43
3353 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3355 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3356 "domain" is actually an IP address, and force a failure. This locks out
3357 those revolvers/nameservers that support "A-for-A" lookups, in
3358 contravention of the specifications.
3360 PH/48 When a host name was looked up from an IP address, and the subsequent
3361 forward lookup of the name timed out, the host name was left in
3362 $sender_host_name, contrary to the specification.
3364 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3365 restricted to single-key lookups, Exim was not diagnosing an error if
3366 * or *@ was used with a query-style lookup.
3368 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3370 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3371 long ago, but noone interested enough thought of it. Let's just say that
3372 the "1.1" means that there are some new functions that weren't there at
3373 some point in the past.
3375 PH/51 Error processing for expansion failure of helo_data from an smtp
3376 transport during callout processing was broken.
3378 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3379 tested/used via the -bh/-bhc/-bs options.
3381 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3382 bug, fixed in subsequent PCRE releases).
3384 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3385 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3387 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3393 SC/01 Use a glob alias rather than an array ref in eximstats generated
3394 parser. This improves both readability and performance.
3396 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3397 Don't display local sender or destination tables in eximstats unless
3398 there is data to show.
3399 Added average volumes into the eximstats top table text output.
3401 SC/03 Collect data on the number of addresses (recipients) as well
3402 as the number of messages in eximstats.
3404 TF/01 Correct an error in the documentation for the redirect router. Exim
3405 does (usually) call initgroups() when daemonizing.
3407 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3408 with consistent privilege compared to when running as a daemon.
3410 TF/03 Note in the spec that $authenticated_id is not set for local
3411 submissions from trusted users.
3413 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3414 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3416 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3417 by adding some example configuration directives to the default
3418 configuration file. A little bit of work is required to uncomment the
3419 directives and define how usernames and passwords are checked, but
3420 there is now a framework to start from.
3422 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3423 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3424 without this. I don't know how relevant this is to other LDAP libraries.
3426 PH/02 Add the verb name to the "unknown ACL verb" error.
3428 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3430 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3432 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3433 directory not expanded when it should be if an expanded home directory
3434 was set for the address (which is overridden by the transport).
3436 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3439 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3440 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3441 because it is too late at that time, and has no effect.
3443 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3444 security issue with \' (bugzilla #107). I could not use the
3445 PQescapeStringConn() function, because it needs a PGconn value as one of
3448 PH/08 When testing addresses using -bt, indicate those final addresses that
3449 are duplicates that would not cause an additional delivery. At least one
3450 person was confused, thinking that -bt output corresponded to deliveries.
3451 (Suppressing duplicates isn't a good idea as you lose the information
3452 about possibly different redirections that led to the duplicates.)
3454 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3455 systems where poll() doesn't work, in particular OS X.
3457 PH/10 Added more information to debugging output for retry time not reached.
3459 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3460 operations in malware.c.
3462 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3465 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3466 syslog_duplication set false, log lines that would normally be written
3467 both the the main log and to the reject log were not written to syslog at
3470 PH/14 In the default configuration, change the use of "message" in ACL warn
3471 statements to "add_header".
3473 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3474 not followed by a command (e.g. "seen endif").
3476 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3477 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3480 PH/17 Added extra conditions to the default value of delay_warning_condition
3484 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3485 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3486 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3489 The Auto-Submitted: and various List- headers are standardised, whereas I
3490 don't think Precedence: ever was.
3492 PH/18 Refactored debugging code in route_finduser() to show more information,
3493 in particular, the error code if getpwnam() issues one.
3495 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3496 This is apparently needed in addition to the PH/07 change above to avoid
3497 any possible encoding problems.
3499 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3500 but not after initializing Perl.
3502 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3503 output them only if debugging. By default they are written stderr,
3504 apparently, which is not desirable.
3506 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3509 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3512 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3514 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3515 authenticated or an ident call has been made. Suppress the default
3516 values for $authenticated_id and $authenticated_sender (but permit -oMai
3517 and -oMas) when testing with -bh.
3519 PH/24 Re-jigged the order of the tests in the default configuration so that the
3520 tests for valid domains and recipients precede the DNS black list and CSA
3521 tests, on the grounds that those ones are more expensive.
3523 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3524 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3525 This bug exists in every version of Exim that I still have, right back to
3528 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3529 However, an attempt to turn on case-sensitivity in a regex key by
3530 including (?-i) didn't work because the subject string was already
3531 lowercased, and the effects were non-intuitive. It turns out that a
3532 one-line patch can be used to allow (?-i) to work as expected.
3538 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3539 other effects) broke the use of negated acl sub-conditions.
3541 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3544 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3545 "Deny" causes Exim to reject the incoming connection with a 554 error.
3546 Unfortunately, if there is a major crisis, such as a disk failure,
3547 tcp-wrappers gives "deny", whereas what one would like would be some
3548 kind of temporary error. A kludge has been added to help with this.
3549 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3550 554 error is used if errno is still zero or contains ENOENT (which occurs
3551 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3554 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3556 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3557 errors. Otherwise a message that provokes a temporary error (when other
3558 messages do not) can cause a whole host to time out.
3560 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3561 addresses were routed directly to files or pipes from a redirect router.
3562 File deliveries just didn't batch; pipe deliveries might have suffered
3565 PH/06 A failure to get a lock for a hints database would erroneously always say
3566 "Failed to get write lock", even when it was really a read lock.
3568 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3569 of 0600. This has been changed to use the value of the lockfile_mode
3570 option (which defaults to 0600).
3572 PH/08 Applied small patch from the Sieve maintainer.
3574 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3575 folder from quota calculations, a direct delivery into this folder messed
3576 up the contents of the maildirsize file. This was because the regex was
3577 used only to exclude .Trash (or whatever) when the size of the mailbox
3578 was calculated. There was no check that a delivery was happening into an
3579 excluded directory. This bug has been fixed by ignoring all quota
3580 processing for deliveries into excluded directories.
3582 PH/10 Added the maildirfolder_create_regex option to appendfile.
3588 PH/01 The code for finding all the local interface addresses on a FreeBSD
3589 system running IPv6 was broken. This may well have applied to all BSD
3590 systems, as well as to others that have similar system calls. The broken
3591 code found IPv4 interfaces correctly, but gave incorrect values for the
3592 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3593 that it would not match correctly against @[] and not recognize the IPv6
3596 PH/02 The ipliteral router was not recognizing addresses of the form user@
3597 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3599 PH/03 Added disable_ipv6.
3601 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3602 decoded form, because it is most often used to construct To: headers
3603 lines in autoreplies, and the decoded form may well be syntactically
3604 invalid. However, $reply_address has leading white space removed, and all
3605 newlines turned into spaces so that the autoreply transport does not
3608 PH/05 If group was specified without a user on a router, and no group or user
3609 was specified on a transport, the group from the router was ignored.
3611 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3612 for visible compile-time settings that can be used to change these
3613 numbers, for those that want even more. Backwards compatibility with old
3614 spool files has been maintained. However, going back to a previous Exim
3615 release will lost any variables that are in spool files.
3617 PH/07 Two small changes when running in the test harness: increase delay when
3618 passing a TCP/IP connection to a new process, in case the original
3619 process has to generate a bounce, and remove special handling of
3620 127.0.0.2 (sic), which is no longer necessary.
3622 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3623 be the same on different OS.
3625 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3628 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3629 whether --show-vars was specified or not
3631 JJ/02 exipick: Added support for new ACL variable spool format introduced
3634 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3635 syntactically invalid From: or Reply-to: line, and a filter used this to
3636 generate an autoreply, and therefore failed to obtain an address for the
3637 autoreply, Exim could try to deliver to a non-existent relative file
3638 name, causing unrelated and misleading errors. What now happens is that
3639 it logs this as a hard delivery error, but does not attempt to create a
3642 PH/11 The exinext utility has a -C option for testing purposes, but although
3643 the given file was scanned by exinext itself; it wasn't being passed on
3644 when Exim was called.
3646 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3647 an end-of-file indication when reading a command response.
3649 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3650 compiled. In many other places in Exim, IPv6 addresses are always
3651 recognized, so I have changed this. It also means that IPv4 domain
3652 literals of the form [IPV4:n.n.n.n] are now always recognized.
3654 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3655 used if the router is not running as root, for example, when verifying at
3656 ACL time, or when using -bh. The debugging output from this situation was
3657 non-existent - all you got was a failure to exec. I have made two
3660 (a) Failures to set uid/gid, the current directory, or a process leader
3661 in a subprocess such as that created by queryprogram now generate
3662 suitable debugging output when -d is set.
3664 (b) The queryprogram router detects when it is not running as root,
3665 outputs suitable debugging information if -d is set, and then runs
3666 the subprocess without attempting to change uid/gid.
3668 PH/15 Minor change to Makefile for building test_host (undocumented testing
3671 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3672 additional section of a DNS packet that returns MX or SRV records.
3673 Instead, it always explicitly searches for A/AAAA records. This avoids
3674 major problems that occur when a DNS server includes only records of one
3675 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3676 fixed another bug: if SRV records were looked up and the corresponding
3677 address records were *not* found in the additional section, the port
3678 values from the SRV records were lost.
3680 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3681 using the correct key (the original address) when searching the retry
3682 rules in order to find which one to use for generating the retry hint.
3684 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3685 adding the default one. Similarly, if it contains a Reply-To: header, the
3686 errors_reply_to option, if set, is not used.
3688 PH/19 When calculating a retry time, Exim used to measure the "time since
3689 failure" by looking at the "first failed" field in the retry record. Now
3690 it does not use this if it is later than than the arrival time of the
3691 message. Instead it uses the arrival time. This makes for better
3692 behaviour in cases where some deliveries succeed, thus re-setting the
3693 "first failed" field. An example is a quota failure for a huge message
3694 when small messages continue to be delivered. Without this change, the
3695 "time since failure" will always be short, possible causing more frequent
3696 delivery attempts for the huge message than are intended.
3697 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3699 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3700 $1, $2, $3) because the numerical variables can be reset during some
3701 expansion items (e.g. "match"), thereby losing the authentication data.
3703 PH/21 Make -bV show the size of off_t variables so that the test suite can
3704 decide whether to run tests for quotas > 2G.
3706 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3707 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3708 filecount value is greater than 2G or if a quota value is greater than 2G
3709 on a system where the size of off_t is not greater than 4, a panic error
3712 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3713 never match. The debug and -bh output now contains an explicit error
3714 message indicating a malformed IPv4 address or mask.
3716 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3717 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3718 PH/23 above applies.
3720 PH/25 Do not write to syslog when running in the test harness. The only
3721 occasion when this arises is a failure to open the main or panic logs
3722 (for which there is an explicit test).
3724 PH/26 Added the /no_tell option to "control=freeze".
3726 PH/27 If a host name lookup failed very early in a connection, for example, if
3727 the IP address matched host_lookup and the reverse lookup yielded a name
3728 that did not have a forward lookup, an error message of the form "no IP
3729 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3730 could be logged. Now it outputs the IP address instead of "NULL".
3732 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3733 allows the sender and the authenticated sender to be set when
3734 submitting a message from within Exim. Since child_open_exim() is
3735 documented for local_scan(), the new function should be too.
3737 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3738 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3739 results in an empty string is now treated as unset.
3741 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3743 PH/31 Added sender_verify_fail logging option.
3745 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3746 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3747 tidied the source and removed it altogether.
3749 PH/33 When a queue run was abandoned because the load average was too high, a
3750 log line was always written; now it is written only if the queue_run log
3751 selector is set. In addition, the log line for abandonment now contains
3752 information about the queue run such as the pid. This is always present
3753 in "start" and "stop" lines but was omitted from the "abandon" line.
3755 PH/34 Omit spaces between a header name and the colon in the error message that
3756 is given when verify = headers_syntax fails (if there are lots of them,
3757 the message gets confusing).
3759 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3760 names, as there are now some PTR records that contain slashes. This check
3761 is only to protect against broken name servers that fall over on strange
3762 characters, so the fact that it applies to all lookups doesn't matter.
3764 PH/36 Now that the new test suite is complete, we can remove some of the
3765 special code in Exim that was needed for the old test suite. For example,
3766 sorting DNS records because real resolvers return them in an arbitrary
3767 order. The new test suite's fake resolver always returns records in the
3770 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3771 bounces) except when queue_only is set, to avoid logging races between
3772 the different processes.
3774 PH/38 Panic-die if .include specifies a non-absolute path.
3776 PH/39 A tweak to the "H" retry rule from its user.
3778 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3779 a label. They prevented compilation on older perls.
3781 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3782 a warning to be raised on newish perls.
3784 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3785 on queue. Changes to match documented behaviour of showing count of
3786 messages matching specified criteria.
3788 PH/40 Changed the default ident timeout from 30s to 5s.
3790 PH/41 Added support for the use of login_cap features, on those BSD systems
3791 that have them, for controlling the resources used by pipe deliveries.
3793 PH/42 The content-scanning code uses fopen() to create files in which to put
3794 message data. Previously it was not paying any attention to the mode of
3795 the files. Exim runs with umask(0) because the rest of the code creates
3796 files with open(), and sets the required mode explicitly. Thus, these
3797 files were ending up world-writeable. This was not a big issue, because,
3798 being within the spool directory, they were not world-accessible. I have
3799 created a function called modefopen, which takes an additional mode
3800 argument. It sets umask(777), creates the file, chmods it to the required
3801 mode, then resets the umask. All the relevant calls to fopen() in the
3802 content scanning code have been changed to use this function.
3804 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3805 to 24 hours. This avoids potential overflow problems when processing G
3806 and H retry rules. I suspect nobody ever tinkers with this value.
3808 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3810 PH/45 When the plaintext authenticator is running as a client, the server's
3811 challenges are checked to ensure they are valid base64 strings. By
3812 default, the authentication attempt is cancelled if an invalid string is
3813 received. Setting client_ignore_invalid_base64 true ignores these errors.
3814 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3815 they are received. Thus, the responses can be made to depend on the
3816 challenges. If an invalid string is ignored, an empty string is placed in
3819 PH/46 Messages that are created by the autoreply transport now contains a
3820 References: header, in accordance with RFCs 2822 and 3834.
3822 PH/47 Added authenticated_sender_force to the smtp transport.
3824 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3826 PH/49 Installed latest patch from the Sieve maintainer.
3828 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3829 was also set, the appendfile transport was unnecessarily scanning a
3830 directory of message files (e.g. for maildir delivery) to find the count
3831 of files (along with the size), even though it did not need this
3832 information. It now does the scan only if it needs to find either the
3833 size of the count of files.
3835 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3837 PH/52 Two bugs concerned with error handling when the smtp transport is
3840 (i) Exim was not creating retry information for temporary errors given
3841 for individual recipients after the DATA command when the smtp transport
3842 was used in LMTP mode. This meant that they could be retried too
3843 frequently, and not timed out correctly.
3845 (ii) Exim was setting the flag that allows error details to be returned
3846 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3847 recipients that were returned after the DATA command.
3849 PH/53 This is related to PH/52, but is more general: for any failing address,
3850 when detailed error information was permitted to be returned to the
3851 sender, but the error was temporary, then after the final timeout, only
3852 "retry timeout exceeded" was returned. Now it returns the full error as
3853 well as "retry timeout exceeded".
3855 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3856 do this, and (what is worse) MTAs that accept it.
3858 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3859 will now be deprecated.
3861 PH/56 New os.c-cygwin from the Cygwin maintainer.
3863 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3864 formats (previously only available in exim formats via -bpr, -bpru,
3865 and -bpra. Now also available in native and exiqgrep formats)
3867 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3868 with very large, slow to parse queues
3870 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3872 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3874 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3875 responses to authentication challenges, though it was showing the
3876 challenges; (ii) I've removed the CR characters from the debug output for
3879 PH/58 Allow for the insertion of a newline as well as a space when a string
3880 is turned into more than one encoded-word during RFC 2047 encoding. The
3881 Sieve code now uses this.
3883 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3884 data_4xx, lost_connection, tls_required.
3886 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3887 message was being sent as an SMTP response.
3889 PH/61 Add -l and -k options to exicyclog.
3891 PH/62 When verifying, if an address was redirected to one new address, so that
3892 verification continued, and the new address failed or deferred after
3893 having set something in $address_data, the value of $address_data was not
3894 passed back to the ACL. This was different to the case when no
3895 redirection occurred. The value is now passed back in both cases.
3897 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3898 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3899 use login_cap.h, so on its own it isn't the distinguishing feature. The
3900 new name refers directly to the setclassresources() function.
3902 PH/65 Added configuration files for NetBSD3.
3904 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3906 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3907 is preferred over IPv4.
3909 PH/68 The bounce_return_message and bounce_return_body options were not being
3910 honoured for bounces generated during the reception of non-SMTP messages.
3911 In particular, this applied to messages rejected by the ACL. This bug has
3912 been fixed. However, if bounce_return_message is true and bounce_return_
3913 body is false, the headers that are returned for a non-SMTP message
3914 include only those that have been read before the error was detected.
3915 (In the case of an ACL rejection, they have all been read.)
3917 PH/69 The HTML version of the specification is now built in a directory called
3918 spec_html instead of spec.html, because the latter looks like a path with
3919 a MIME-type, and this confuses some software.
3921 PH/70 Catch two compiler warnings in sieve.c.
3923 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3924 function verify_get_ident() calls ip_connect() to connect a socket, but
3925 if the "connect()" function timed out, ip_connect() used to close the
3926 socket. However, verify_get_ident() also closes the socket later, and in
3927 between Exim writes to the log, which may get opened at this point. When
3928 the socket was closed in ip_connect(), the log could get the same file
3929 descriptor number as the socket. This naturally causes chaos. The fix is
3930 not to close the socket in ip_connect(); the socket should be closed by
3931 the function that creates it. There was only one place in the code where
3932 this was missing, in the iplookup router, which I don't think anybody now
3933 uses, but I've fixed it anyway.
3935 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3936 well as to direct DNS lookups. Otherwise the handling of names in host
3937 lists is inconsistent and therefore confusing.
3943 PH/01 Two changes to the default runtime configuration:
3945 (1) Move the checks for relay_from_hosts and authenticated clients from
3946 after to before the (commented out) DNS black list checks.
3948 (2) Add control=submission to the relay_from_hosts and authenticated
3949 clients checks, on the grounds that messages accepted by these
3950 statements are most likely to be submissions.
3952 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3954 (1) Generate an error if the third argument for the ${prvs expansion is
3957 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3960 (3) Reset the variables that are obtained from the first argument of
3961 ${prvscheck and used in the second argument before leaving the code,
3962 because their memory is reclaimed, so using them afterwards may do
3965 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3966 one (it's much easier than Tom thought :-).
3968 (5) Because of (4), we can now allow for the use of $prvscheck_result
3969 inside the third argument.
3971 PH/03 For some reason, the default setting of PATH when running a command from
3972 a pipe transport was just "/usr/bin". I have changed it to
3975 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3976 anything to be listed in the output from -bV.
3978 PH/05 When a filter generated an autoreply, the entire To: header line was
3979 quoted in the delivery log line, like this:
3981 => >A.N.Other <ano@some.domain> <original@ddress> ...
3983 This has been changed so that it extracts the operative address. There
3984 may be more than one such address. If so, they are comma-separated, like
3987 => >ano@some.domain,ona@other.domain <original@ddress> ...
3989 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3990 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3991 not being looked up in the rDNS to get a host name, Exim was showing the
3992 IP address twice in Received: lines, even though the IP addresses were
3993 identical. For example:
3995 Received: from [1.2.3.4] (helo=[1.2.3.4])
3997 However, if the real host name was known, it was omitting the HELO data
3998 if it matched the actual IP address. This has been tidied up so that it
3999 doesn't show the same IP address twice.
4001 PH/07 When both +timestamp and +memory debugging was on, the value given by
4002 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4003 called by the debug printing, thereby overwriting the timestamp buffer.
4004 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4006 PH/08 When the original message was included in an autoreply transport, it
4007 always said "this is a copy of the message, including all the headers",
4008 even if body_only or headers_only was set. It now gives an appropriate
4011 PH/09 Applied a patch from the Sieve maintainer which:
4013 o fixes some comments
4014 o adds the (disabled) notify extension core
4015 o adds some debug output for the result of if/elsif tests
4016 o points to the current vacation draft in the documentation
4017 and documents the missing references header update
4021 o fixes a bug in processing the envelope test (when testing
4022 multiple envelope elements, the last element determined the
4025 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4026 Electronic Mail") by including:
4028 Auto-submitted: auto-generated
4030 in the messages that it generates (bounce messages and others, such as
4031 warnings). In the case of bounce messages for non-SMTP messages, there was
4032 also a typo: it was using "Auto_submitted" (underscore instead of
4033 hyphen). Since every message generated by Exim is necessarily in response
4034 to another message, thes have all been changed to:
4036 Auto-Submitted: auto-replied
4038 in accordance with these statements in the RFC:
4040 The auto-replied keyword:
4042 - SHOULD be used on messages sent in direct response to another
4043 message by an automatic process,
4045 - MUST NOT be used on manually-generated messages,
4047 - MAY be used on Delivery Status Notifications (DSNs) and Message
4048 Disposition Notifications (MDNs),
4050 - MUST NOT be used on messages generated by automatic or periodic
4051 processes, except for messages which are automatic responses to
4054 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4055 to the default Received: header definition.
4057 PH/12 Added log selector acl_warn_skipped (default on).
4059 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4060 variables because (a) they are in the wrong storage pool and (b) even if
4061 they were copied, it wouldn't work properly because of the caching.
4063 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4064 checking when decoding. Apparently there are clients that generate
4065 overlong encoded strings. Why am I not surprised?
4067 PH/15 If the first argument of "${if match_address" was not empty, but did not
4068 contain an "@" character, Exim crashed. Now it writes a panic log message
4069 and treats the condition as false.
4071 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4073 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4074 extension "envelope-auth". The code is finished and in agreement with
4075 other implementations, but there is no documentation so far and in fact,
4076 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4077 not changing the active code.
4079 Print executed "if" and "elsif" statements when debugging is used. This
4080 helps a great deal to understand what a filter does.
4082 Document more things not specified clearly in RFC3028. I had all this
4083 sorted out, when out of a sudden new issues came to my mind. Oops."
4085 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4088 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4089 canonical form (as documented). However, after a host name lookup from
4090 the IP address, check_host() was doing a simple string comparison with
4091 addresses acquired from the DNS when checking that the found name did
4092 have the original IP as one of its addresses. Since any found IPv6
4093 addresses are likely to be in abbreviated form, the comparison could
4094 fail. Luckily, there already exists a function for doing the comparison
4095 by converting both addresses to binary, so now that is used instead of
4096 the text comparison.
4098 PH/20 There was another similar case to PH/19, when a complete host name was
4099 given in a host list; looking up its IP address could give an abbreviated
4100 form, whereas the current host's name might or might not be abbreviated.
4101 The same fix has been applied.
4107 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4108 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4111 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4112 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4114 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4116 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4117 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4118 Also, for "Auto-Submitted", treat anything other than "no" as
4119 non-personal, in accordance with RFC 3834. (Previously it treated
4120 anything starting "auto-" as non-personal.)
4122 TF/01 The control=submission/name=... option had a problem with syntax
4123 errors if the name included a slash character. The /name= option
4124 now slurps the rest of the string, so it can include any characters
4125 but it must come last in the list of options (after /sender_retain
4128 PH/05 Some modifications to the interface to the fake nameserver for the new
4136 TK/01 Added the "success_on_redirect" address verification option. See
4137 NewStuff for rationale and an example.
4139 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4141 PH/02 Patch to exigrep to allow it to work on syslog lines.
4143 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4144 fread() to skip over the body file's header line, because in Cygwin the
4145 header line is locked and is inaccessible.
4147 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4148 co-exist for some time) to make it clear that it is the Exim ID that is
4149 referenced, not the Message-ID: header line.
4151 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4152 string_format() function, because snprintf() does not exist on all
4155 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4156 ${stat: expansion item.
4158 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4159 protocol synchronization error", to keep the pedants happy.
4161 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4162 well as for IRIX systems, when gcc is being used. See the host.c source
4165 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4167 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4170 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4171 transport and to the smtp transport in LMTP mode.
4173 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4175 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4176 run a filter in a subprocess. This could lead to confusion in subsequent
4177 lookups in the parent process. There should also be a search_tidyup() at
4178 the end of the subprocess.
4180 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4181 only if the host matched helo_try_verify_hosts, which caused the
4182 verification to occur when the EHLO/HELO command was issued. The ACL just
4183 tested the remembered result. Now, if a previous verification attempt has
4184 not happened, "verify = helo" does it there and then.
4186 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4188 TK/03 Fix log output including CR from clamd.
4190 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4191 exist provoked a memory error which could cause a segfault.
4193 PH/15 Installed PCRE 6.2
4195 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4197 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4198 of the problem. Specifically, suggested +O2 rather than +O1 for the
4201 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4203 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4204 record, the port was not correctly logged when the outgoing_port log
4205 selector was set (it logged the transort's default port).
4207 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4208 fallback_hosts, and "hosts" in the smtp transport.
4210 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4211 host errors such as "Connection refused".
4213 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4214 authentication with radiusclient 0.4.9:
4216 - Error returned from rc_read_config was caught wrongly
4217 - Username/password not passed on to radius server due to wrong length.
4219 The presumption is that some radiusclient API changes for 4.51/PH/17
4220 were not taken care of correctly. The code is still untested by me (my
4221 Linux distribution still has 0.3.2 of radiusclient), but it was
4222 contributed by a Radius user.
4224 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4225 expanding the "port" option of the smtp transport.
4227 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4228 while reading a MIME header. Thanks to Tom Hughes for a patch.
4230 PH/24 Include config.h inside local_scan.h so that configuration settings are
4233 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4234 that in an ACL for RCPT (for example), you can examine exactly what was
4237 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4238 commands, but it was not correctly comparing the address with the actual
4239 client host address. Thus, it would show the EHLO address in Received:
4240 header lines when this was not necessary.
4242 PH/27 Added the % operator to ${eval:}.
4244 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4245 it should be ignoring failures (because with -C, for example, it has lost
4246 privilege). It wasn't ignoring creation failures other than "already
4249 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4252 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4253 because an input line was too long, either on its own, or by virtue of
4254 too many continuations, the temporary file was not being removed, and the
4255 return code was incorrect.
4257 PH/31 Missing "BOOL" in function definition in filtertest.c.
4259 PH/32 Applied Sieve patches from the maintainer.
4261 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4263 PH/33 Added "verify = not_blind".
4265 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4266 Local/Makefile (with some defaults set). These are used in built scripts
4267 such as exicyclog, but they have never been used in the exim_install
4268 script (though there are many overriding facilities there). I have
4269 arranged that the exim_install script now takes note of these two
4272 PH/35 Installed configuration files for Dragonfly.
4274 PH/36 When a locally submitted message by a trusted user did not contain a
4275 From: header, and the sender address was obtained from -f or from an SMTP
4276 MAIL command, and the trusted user did not use -F to supply a sender
4277 name, $originator_name was incorrectly used when constructing a From:
4278 header. Furthermore, $originator_name was used for submission mode
4279 messages from external hosts without From: headers in a similar way,
4280 which is clearly wrong.
4282 PH/37 Added control=suppress_local_fixups.
4284 PH/38 When log_selector = +received_sender was set, and the addition of the
4285 sender made the log line's construction buffer exactly full, or one byte
4286 less than full, an overflow happened when the terminating "\n" was
4289 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4290 when the result of a list match is failure because a DNS lookup failed.
4292 PH/40 RM_COMMAND is now used in the building process.
4294 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4295 the "build-* directories that it finds.
4297 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4298 address in a domain literal was a prefix of an interface address.
4300 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4301 when verifying a sender address, unless rewrite_headers is false.
4303 PH/44 Wrote a long comment about why errors_to addresses are verified as
4304 recipients, not senders.
4306 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4307 the ratelimit ACL was added.
4309 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4311 PH/46 Added extra information about PostgreSQL errors to the error string.
4313 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4314 avoiding the need to install special zones in a real server. This is
4315 backwards compatible; if it can't find the fake resolver, it drops back.
4316 Thus, both old and new test suites can be run.
4318 TF/02 Added util/ratelimit.pl
4320 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4321 clock is set back in time.
4323 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4324 Candler <B.Candler@pobox.com>.
4326 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4327 OK for addresses that are the result of redirections.
4329 PH/48 A number of further additions for the benefit of the new test suite,
4330 including a fake gethostbyname() that interfaces to the fake DNS resolver
4333 TF/06 The fix for widen_domains has also been applied to qualify_single and
4334 search_parents which are the other dnslookup options that can cause
4337 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4340 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4342 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4343 service" if it encountered an MX record with an empty target hostname.
4344 The message is now "an MX or SRV record indicated no SMTP service".
4346 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4347 if the DNS of the sending site is misconfigured. This is quite a
4348 common situation. This change restores the behaviour of treating a
4349 helo verification defer as a failure.
4351 PH/51 If self=fail was set on a router, the bounce message did not include the
4352 actual error message.
4358 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4360 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4361 command itself ended in error, the underlying message about the transport
4362 filter timeout was being overwritten with the pipe command error. Now the
4363 underlying error message should be appended to the second error message.
4365 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4367 PH/02 Reduce the amount of output that "make" produces by default. Full output
4368 can still be requested.
4370 PH/03 The warning log line about a condition test deferring for a "warn" verb
4371 was being output only once per connection, rather than after each
4372 occurrence (because it was using the same function as for successful
4373 "warn" verbs). This seems wrong, so I have changed it.
4375 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4376 it should not have, which might have caused a crash in the right
4377 circumstances, but probably never did.
4379 PH/04 Installed a modified version of Tony Finch's patch to make submission
4380 mode fix the return path as well as the Sender: header line, and to
4381 add a /name= option so that you can make the user's friendly name appear
4384 TF/03 Added the control = fakedefer ACL modifier.
4386 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4387 Mark Lowes for thorough testing.
4389 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4391 TK/03 Merged latest SRS patch from Miles Wilton.
4393 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4394 with the definition in sysexits.h (which is #included earlier).
4395 Fortunately, Exim does not actually use EX_OK. The code used to try to
4396 preserve the sysexits.h value, by assuming that macro definitions were
4397 scanned for macro replacements. I have been disabused of this notion,
4398 so now the code just undefines EX_OK before #including unistd.h.
4400 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4401 in the smtp transport. When a block could not be written in a single
4402 write() function, the timeout was being re-applied to each part-write.
4403 This seems wrong - if the receiver was accepting one byte at a time it
4404 would take for ever. The timeout is now adjusted when this happens. It
4405 doesn't have to be particularly precise.
4407 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4408 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4410 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4411 without a domain if the check to <postmaster@domain> fails.
4413 SC/01 Eximstats: added -xls and the ability to specify output files
4414 (patch written by Frank Heydlauf).
4416 SC/02 Eximstats: use FileHandles for outputting results.
4418 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4420 SC/04 Eximstats: fixed display of large numbers with -nvr option
4422 SC/05 Eximstats: fixed merging of reports with empty tables.
4424 SC/06 Eximstats: added the -include_original_destination flag
4426 SC/07 Eximstats: removed tabs and trailing whitespace.
4428 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4430 TK/06 MBOX spool code: Add real "From " MBOX separator line
4431 so the .eml file is really in mbox format (even though
4432 most programs do not really care). Patch from Alex Miller.
4434 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4435 The latter is generated from $received_to and is only set if the
4436 message has one envelope recipient. SA can use these headers,
4437 obviously out-of-the-box. Patch from Alex Miller.
4439 PH/08 The ${def test on a variable was returning false if the variable's
4440 value was "0", contrary to what the specification has always said!
4441 The result should be true unless the variable is empty.
4443 PH/09 The syntax error of a character other than { following "${if
4444 def:variable_name" (after optional whitespace) was not being diagnosed.
4445 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4446 accidental colon was present, for example, could give incorrect results.
4448 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4449 result is used (not including appendfile, where other changes are about
4452 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4453 This involved changing a lot of size variables from int to off_t. It
4454 should work with maildirs and everything.
4456 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4457 spamd dying while we are connected to it.
4459 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4462 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4463 function for BDB 4.3.
4465 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4467 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4468 the sources for PCRE have been re-arranged and more files are now
4471 PH/15 The code I had for printing potentially long long variables in PH/11
4472 above was not the best (it lost precision). The length of off_t variables
4473 is now inspected at build time, and an appropriate printing format (%ld
4474 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4475 to be "long long int" or "long int". This is needed for the internal
4476 formatting function string_vformat().
4478 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4479 the configuration file to be ":syslog", then the script "guesses" where
4480 the logs files are, rather than using the compiled in default. In our
4481 case the guess is not the same as the compiled default, so the script
4482 suddenly stopped working when I started to use syslog. The patch checks
4483 to see if log_file_path is "". If so, it attempts to read it from exim
4484 with no configuration file to get the compiled in version, before it
4485 falls back to the previous guessing code."
4487 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4488 implementing BATV in an Exim configuration. See NewStuff for the gory
4491 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4492 Makefile that are specific to HP-UX.
4494 PH/18 If the "use_postmaster" option was set for a recipient callout together
4495 with the "random" option, the postmaster address was used as the MAIL
4496 FROM address for the random test, but not for the subsequent recipient
4497 test. It is now used for both.
4499 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4500 patch removes a few documentation additions to RFC 3028, because the
4501 latest draft now contains them. It adds the new en;ascii-case comparator
4502 and a new error check for 8bit text in MIME parts. Comparator and
4503 require names are now matched exactly. I enabled the subaddress
4504 extension, but it is not well tested yet (read: it works for me)."
4506 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4507 rework some of the code of TK/09 above to avoid the hardwired use of
4508 "%lld" and "long long". Replaced the call to snprintf() with a call to
4511 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4512 records point to non-existent hosts", "retry timeout exceeded", and
4513 "retry time not reached for any host after a long failure period".
4515 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4516 experimental DomainKeys support:
4518 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4519 (2) On an error such as an illegally used "control", the wrong name for
4520 the control was given.
4522 These problems did NOT occur unless DomainKeys support was compiled.
4524 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4526 PH/24 Added ${if match_ip condition.
4528 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4529 databases so that it will be absolutely obvious if a crash occurs in the
4530 DB library. This is a regular occurrence (often caused by mis-matched
4533 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4534 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4535 picked up on a user's system that detects such things. There doesn't seem
4536 to be a gcc warning option for this - only an attribute that has to be
4537 put on the function's prototype. It seems that in Fedora Core 4 they have
4538 set this on a number of new functions. No doubt there will be more in due
4541 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4542 specify a transport. However, if an address that was verified by such a
4543 router was the subject of a callout, Exim crashed because it tried to
4544 read the rcpt_include_affixes from the non-existent transport. Now it
4545 just assumes that the setting of that option is false. This bug was
4546 introduced by 4.51/PH/31.
4548 PH/28 Changed -d+all to exclude +memory, because that information is very
4549 rarely of interest, but it makes the output a lot bigger. People tend to
4550 do -d+all out of habit.
4552 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4553 code in os-type was giving problems when libc.so lives in lib64, like on
4556 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4557 aren't the modern standard, and it seems that some systems' include files
4558 don't always have them. Exim was already checking for some of the newer
4559 ones like T_AAAA, and defining it itself. I've added checks for all the
4560 record types that Exim uses.
4562 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4563 not automatically generating a new one, as it is supposed to. This
4564 prevented TLS from working. If the file did exist, but contained invalid
4565 data, a new version was generated, as expected. It was only the case of a
4566 non-existent file that was broken.
4568 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4569 with a change in libdomainkeys > 0.64.
4571 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4572 from DNS. If the selector record carries the flag, it now has
4573 precedence over the domain-wide flag.
4575 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4577 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4578 the use of a port name that isn't defined in /etc/services) occurred, the
4579 message was deferred as in a normal delivery, and thus remained on the
4580 spool, instead of being failed because of the mua_wrapper setting. This
4581 is now fixed, and I tidied up some of the mua_wrapper messages at the
4584 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4585 the messages in a hash of arrays rather than using individual hashes.
4586 This is a bit cleaner and results in dramatic memory savings, albeit
4587 at a slight CPU cost.
4589 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4590 as requested by Marc Sherman.
4592 SC/10 Eximstats: added histograms for user specified patterns as requested
4595 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4597 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4598 fopen() in the content-scanning modules that did not already have it.
4604 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4605 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4607 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4609 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4611 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4612 to test Sieve filters that use "vacation".
4614 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4615 that changes the way the GnuTLS parameters are stored in the cache file.
4616 The new format can be generated externally. For backward compatibility,
4617 if the data in the cache doesn't make sense, Exim assumes it has read an
4618 old-format file, and it generates new data and writes a new file. This
4619 means that you can't go back to an older release without removing the
4622 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4623 work if there are any delivery delays because "one_time" forces the
4624 parent to be marked "delivered", so its unseen clone is never tried
4625 again. For this reason, Exim now forbids the simultaneous setting of
4628 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4629 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4630 entry if you want to know the details. The fix, however, neglected to
4631 consider the case when local delivery batching is involved. The test for
4632 "previously delivered" was not happening when checking to see if an
4633 address could be batched with a previous (undelivered) one; under
4634 certain circumstances this could lead to multiple deliveries to the same
4637 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4638 in its include files, and this causes problems building Exim.
4640 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4641 header_syntax) but Exim was just ignoring anything given after a slash.
4642 In particular, this caused confusion with an attempt to use "verify =
4643 reverse_host_lookup/defer_ok". An error is now given when options are
4644 supplied for verify items that do not have them. (Maybe reverse_host_
4645 lookup should have a defer_ok option, but that's a different point.)
4647 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4648 defined by RFC 821) to 2048, because there were problems with some AUTH
4649 commands, and RFC 1869 says the size should be increased for extended
4650 SMTP commands that take arguments.
4652 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4655 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4656 "unknown" error; now it says that the functionality isn't in the binary.
4658 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4659 an address' error message when a string expansion fails (syntax or
4660 whatever). Otherwise the password may appear in the log. Following change
4661 PH/42 below, there is no longer a chance of it appearing in a bounce
4664 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4666 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4667 of its addresses was ever tried. (Bugzilla bug #2.)
4669 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4670 the result incorrectly in the debug output. (It correctly added a newline
4671 to what was transported.)
4673 TF/01 Added $received_time.
4675 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4676 commented out examples of how to interface to a virus scanner and to
4677 SpamAssassin. Also added commented examples of av_scanner and
4678 spamd_address settings.
4680 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4681 and controls are allowed in which ACLs. There were a couple of minor
4682 errors. Some of the entries in the conditions table (which is a table of
4683 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4684 negation of where the condition IS allowed.
4686 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4688 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4689 header file does not have a version number, so I've had to invent a new
4690 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4691 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4692 radiusclient), but it was contributed by a Radius user.
4694 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4695 files or directories, for OpenSSL.
4697 PH/19 When an Exim process that is running as root has to create an Exim log
4698 file, it does so in a subprocess that runs as exim:exim so as to get the
4699 ownership right at creation (otherwise, other Exim processes might see
4700 the file with the wrong ownership). There was no test for failure of this
4701 fork() call, which would lead to the process getting stuck as it waited
4702 for a non-existent subprocess. Forks do occasionally fail when resources
4703 run out. I reviewed all the other calls to fork(); they all seem to check
4706 PH/20 When checking for unexpected SMTP input at connect time (before writing
4707 the banner), Exim was not dealing correctly with a non-positive return
4708 from the read() function. If the client had disconnected by this time,
4709 the result was a log entry for a synchronization error with an empty
4710 string after "input=" when read() returned zero. If read() returned -1
4711 (an event I could not check), uninitialized data bytes were printed.
4712 There were reports of junk text (parts of files, etc) appearing after
4715 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4717 PH/22 Added support for macro redefinition, and (re)definition in between
4718 driver and ACL definitions.
4720 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4721 forgetting to use the resulting value; it was using the unexpanded value.
4723 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4724 hadn't been configured. The fix is from Juergen Kreileder, who
4725 understands it better than I do:
4727 "Here's what I see happening with three configured cyrus_sasl
4728 authenticators configured (plain, login, cram-md5):
4730 On startup auth_cyrus_sasl_init() gets called for each of these.
4731 This means three calls to sasl_listmech() without a specified mech_list.
4732 => SASL tests which mechs of all available mechs actually work
4733 => three warnings about OTP not working
4734 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4736 With the patch, sasl_listmech() also gets called three times. But now
4737 SASL's mech_list option is set to the server_mech specified in the the
4738 authenticator. Or in other words, the answer from sasl_listmech()
4739 gets limited to just the mech you're testing for (which is different
4741 => the return list contains just 'plain' or 'login', 'cram-md5' or
4742 nothing depending on the value of ob->server_mech.
4744 I've just tested the patch: Authentication still works fine,
4745 unavailable mechs specified in the exim configuration are still
4746 caught, and the auth.log warnings about OTP are gone."
4748 PH/25 When debugging is enabled, the contents of the command line are added
4749 to the debugging output, even when log_selector=+arguments is not
4752 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4753 answer is "GNU", and only if the return is "GNU/something" is the answer
4756 PH/27 $acl_verify_message is now set immediately after the failure of a
4757 verification in an ACL, and so is available in subsequent modifiers. In
4758 particular, the message can be preserved by coding like this:
4760 warn !verify = sender
4761 set acl_m0 = $acl_verify_message
4763 Previously, $acl_verify_message was set only while expanding "message"
4764 and "log_message" when a very denied access.
4766 PH/28 Modified OS/os.c-Linux with
4768 -#ifndef OS_LOAD_AVERAGE
4769 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4771 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4772 nomenclature these days.)
4774 PH/29 Installed patch from the Sieve maintainer that adds the options
4775 sieve_useraddress and sieve_subaddress to the redirect router.
4777 PH/30 In these circumstances:
4778 . Two addresses routed to the same list of hosts;
4779 . First host does not offer TLS;
4780 . First host accepts first address;
4781 . First host gives temporary error to second address;
4782 . Second host offers TLS and a TLS session is established;
4783 . Second host accepts second address.
4784 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4785 and peerdn, if requested) that were in fact used only for the second
4788 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4789 attention to any local part prefix or suffix that was matched by the
4790 router that accepted the address. It now behaves in the same way as it
4791 does for delivery: the affixes are removed from the local part unless
4792 rcpt_include_affixes is set on the transport.
4794 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4795 timeout during the DATA phase of an incoming message.
4797 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4798 applied a patch sanctioned by the Sieve maintainer.
4800 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4801 the uid or gid is negative. A case of a negative gid caused this to be
4802 noticed. The fix allows for either to be negative.
4804 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4805 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4808 PH/36 The change PH/12 above was broken. Fixed it.
4810 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4811 the grounds that routing the same address twice would always produce the
4812 same answer. This might have been true once, but it is certainly no
4813 longer true now. Routing a child address may depend on the previous
4814 routing that produced that child. Some complicated redirection strategies
4815 went wrong when messages had multiple recipients, and made Exim's
4816 behaviour dependent on the order in which the addresses were given.
4818 I have moved the duplicate checking until after the routing is complete.
4819 Exim scans the addresses that are assigned to local and remote
4820 transports, and removes any duplicates. This means that more work will be
4821 done, as duplicates will always all be routed, but duplicates are
4822 presumably rare, so I don't expect this is of any significance.
4824 For deliveries to pipes, files, and autoreplies, the duplicate checking
4825 still happens during the routing process, since they are not going to be
4828 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4829 It corrects a timeout issue with spamd. This is Ian's comment: "The
4830 background is that sometimes spamd either never reads data from a
4831 connection it has accepted, or it never writes response data. The exiscan
4832 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4833 blindly assumes that writes won't block so it may never time out."
4835 PH/39 Allow G after quota size as well as K and M.
4837 PH/40 The value set for $authenticated_id in an authenticator may not contain
4838 binary zeroes or newlines because the value is written to log lines and
4839 to spool files. There was no check on this. Now the value is run through
4840 the string_printing() function so that such characters are converted to
4841 printable escape sequences.
4843 PH/41 $message_linecount is a new variable that contains the total number of
4844 lines in the message. Compare $body_linecount, which is the count for the
4847 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4848 bounce and delay warning messages, except in certain special cases, which
4851 (a) An SMTP error message from a remote host;
4852 (b) A message specified in a :fail: redirection;
4853 (c) A message specified in a "fail" command in a system filter;
4854 (d) A message specified in a FAIL return from the queryprogram router;
4855 (e) A message specified by the cannot_route_message router option.
4857 In these cases only, Exim does include the error details in bounce and
4858 warning messages. There are also a few cases where bland messages such
4859 as "unrouteable address" or "local delivery error" are given.
4861 PH/43 $value is now also set for the "else" part of a ${run expansion.
4863 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4864 being worked on, but at least Exim now implements the latest version to
4867 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4868 process to complete was treated as a delivery failure, a timeout while
4869 writing the message to the pipe was logged, but erroneously treated as a
4870 successful delivery. Such timeouts include transport filter timeouts. For
4871 consistency with the overall process timeout, these timeouts are now
4872 treated as errors, giving rise to delivery failures by default. However,
4873 there is now a new Boolean option for the pipe transport called
4874 timeout_defer, which, if set TRUE, converts the failures into defers for
4875 both kinds of timeout. A transport filter timeout is now identified in
4878 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4879 systems where "make" and "gmake" are different, calling "gmake" at top
4880 level broke things. I've arranged for the value of $(MAKE) to be passed
4881 from the Makefile to this script so that it can call the same version of
4885 A note about Exim versions 4.44 and 4.50
4886 ----------------------------------------
4888 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4889 changes of various kinds. As a consequence, a big documentation update was
4890 needed. This delayed the release for rather longer than seemed good, especially
4891 in the light of a couple of (minor) security issues. Therefore, the changes
4892 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4893 release. So 4.44 and 4.50 are in effect two different branches that both start
4896 I have left the 4.50 change log unchanged; it contains all the changes since
4897 4.43. The change log for 4.44 is below; many of its items are identical to
4898 those for 4.50. This seems to be the most sensible way to preserve the
4899 historical information.
4905 1. Minor wording change to the doc/README.SIEVE file.
4907 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4908 computation of the current number of files was incorrect.
4910 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4911 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4914 4. Give more explanation in the error message when the command for a transport
4915 filter fails to execute.
4917 5. There are several places where Exim runs a non-Exim command in a
4918 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4919 was being done only for the command run by the queryprogram router. It is
4920 now done for all such subprocesses. The other cases are: ${run, transport
4921 filters, and the commands run by the lmtp and pipe transports.
4923 6. Added CONFIGURE_GROUP build-time option.
4925 7. Some older OS have a limit of 256 on the maximum number of file
4926 descriptors. Exim was using setrlimit() to set 1000 as a large value
4927 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4928 systems. I've change it so that if it can't get 1000, it tries for 256.
4930 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4931 was an oversight, and furthermore, ever since the addition of extra
4932 controls (e.g. 4.43/32), the checks on when to allow different forms of
4933 "control" were broken. There should now be diagnostics for all cases when a
4934 control that does not make sense is encountered.
4936 9. Added the /retain_sender option to "control=submission".
4938 10. $recipients is now available in the predata ACL (oversight).
4940 11. Tidy the search cache before the fork to do a delivery from a message
4941 received from the command line. Otherwise the child will trigger a lookup
4942 failure and thereby defer the delivery if it tries to use (for example) a
4943 cached ldap connection that the parent has called unbind on.
4945 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4946 of $address_data from the recipient verification was clobbered by the
4947 sender verification.
4949 13. The value of address_data from a sender verification is now available in
4950 $sender_address_data in subsequent conditions in the ACL statement.
4952 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4954 15. Added a new option "connect=<time>" to callout options, to set a different
4957 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4958 was its contents. (It was OK if the option was not defined at all.)
4960 17. A "Completed" log line is now written for messages that are removed from
4961 the spool by the -Mrm option.
4963 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4964 information about exactly what failed.
4966 19. Added -dd to debug only the daemon process.
4968 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4969 handles timeouts, both on the server side and network timeouts. Renamed the
4970 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4972 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4973 It is now set to "smtps".
4975 22. $host_address is now set to the target address during the checking of
4976 ignore_target_hosts.
4978 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4979 being passed; this would have caused $sender_host_name to have been used if
4980 matching the list had actually called for a host name (not very likely,
4981 since this list is usually IP addresses). A host name is now passed as
4984 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4985 code that specifies a non-restarting handler (typically sigaction() in
4986 modern systems) in an attempt to fix a rare and obscure crash bug.
4988 25. Narrowed the window for a race in the daemon that could cause it to ignore
4989 SIGCHLD signals. This is not a major problem, because they are used only to
4990 wake it up if nothing else does.
4992 26. A malformed maildirsize file could cause Exim to calculate negative values
4993 for the mailbox size or file count. Odd effects could occur as a result.
4994 The maildirsize information is now recalculated if the size or filecount
4997 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4998 support for a long time. Removed HAVE_SYS_VFS_H.
5000 28. Installed the latest version of exipick from John Jetmore.
5002 29. In an address list, if the pattern was not a regular expression, an empty
5003 subject address (from a bounce message) matched only if the pattern was an
5004 empty string. Non-empty patterns were not even tested. This was the wrong
5005 because it is perfectly reasonable to use an empty address as part of a
5006 database query. An empty address is now tested by patterns that are
5007 lookups. However, all the other forms of pattern expect the subject to
5008 contain a local part and a domain, and therefore, for them, an empty
5009 address still always fails if the pattern is not itself empty.
5011 30. Exim went into a mad DNS loop when attempting to do a callout where the
5012 host was specified on an smtp transport, and looking it up yielded more
5013 than one IP address.
5015 31. Re-factored the code for checking spool and log partition space into a
5016 function that finds that data and another that does the check. The former
5017 is then used to implement four new variables: $spool_space, $log_space,
5018 $spool_inodes, and $log_inodes.
5020 32. The RFC2047 encoding function was originally intended for short strings
5021 such as real names; it was not keeping to the 75-character limit for
5022 encoded words that the RFC imposes. It now respects the limit, and
5023 generates multiple encoded words if necessary. To be on the safe side, I
5024 have increased the buffer size for the ${rfc2047: expansion operator from
5027 33. It is now permitted to omit both strings after an "if" condition; if the
5028 condition is true, the result is "true". As before, when the second string
5029 is omitted, a false condition yields an empty string. This makes it less
5030 cumbersome to write custom ACL and router conditions.
5032 34. Failure to deliver a bounce message always caused it to be frozen, even if
5033 there was an errors_to setting on the router. The errors_to setting is now
5036 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5037 canonical form (fully expanded) before being placed in
5038 $sender_host_address.
5040 36. The table in the code that translates DNS record types into text (T_A to
5041 "A" for instance) was missing entries for NS and CNAME. It is just possible
5042 that this could have caused confusion if both these types were looked up
5043 for the same domain, because the text type is used as part of Exim's
5044 per-process caching. But the chance of anyone hitting this buglet seems
5047 37. The dnsdb lookup has been extended in a number of ways.
5049 (1) There is a new type, "zns", which walks up the domain tree until it
5050 finds some nameserver records. It should be used with care.
5052 (2) There is a new type, "mxh", which is like "mx" except that it returns
5053 just the host names, not the priorities.
5055 (3) It is now possible to give a list of domains (or IP addresses) to be
5056 looked up. The behaviour when one of the lookups defers can be
5057 controlled by a keyword.
5059 (4) It is now possible to specify the separator character for use when
5060 multiple records are returned.
5062 38. The dnslists ACL condition has been extended: it is now possible to supply
5063 a list of IP addresses and/or domains to be looked up in a particular DNS
5066 39. Added log_selector=+queue_time_overall.
5068 40. When running the queue in the test harness, wait just a tad after forking a
5069 delivery process, to get repeatability of debugging output.
5071 41. Include certificate and key file names in error message when GnuTLS fails
5072 to set them up, because the GnuTLS error message doesn't include the name
5073 of the failing file when there is a problem reading it.
5075 42. Allow both -bf and -bF in the same test run.
5077 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5079 44. The "Exiscan patch" is now merged into the mainline Exim source.
5081 45. Sometimes the final signoff response after QUIT could fail to get
5082 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5083 < 0 before doing a fflush(). This bug looks as though it goes back to the
5084 introduction of TLS in release 3.20, but "sometimes" must have been rare
5085 because the tests only now provoked it.
5087 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5088 (this can affect the format of dates).
5090 47. exim_tidydb, when checking for the continued existence of a message for
5091 which it has found a message-specific retry record, was not finding
5092 messages that were in split spool directories. Consequently, it was
5093 deleting retry records that should have stayed in existence.
5095 48. Steve fixed some bugs in eximstats.
5097 49. The SPA authentication driver was not abandoning authentication and moving
5098 on to the next authenticator when an expansion was forced to fail,
5099 contradicting the general specification for all authenticators. Instead it
5100 was generating a temporary error. It now behaves as specified.
5102 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5103 (the order specifies the preference for clients). The order is now AES256,
5104 AES128, 3DES, ARCFOUR128.
5106 51. Small patch to Sieve code - explicitly set From: when generating an
5109 52. Exim crashed if a remote delivery caused a very long error message to be
5110 recorded - for instance if somebody sent an entire SpamAssassin report back
5111 as a large number of 550 error lines. This bug was coincidentally fixed by
5112 increasing the size of one of Exim's internal buffers (big_buffer) that
5113 happened as part of the Exiscan merge. However, to be on the safe side, I
5114 have made the code more robust (and fixed the comments that describe what
5117 53. Now that there can be additional text after "Completed" in log lines (if
5118 the queue_time_overall log selector is set), a one-byte patch to exigrep
5119 was needed to allow it to recognize "Completed" as not the last thing in
5122 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5123 patch that reportedly fixes this has been added. I am not expert enough to
5124 create a test for it. This is what the patch creator wrote:
5126 "I found a little strange behaviour of ldap code when working with
5127 Windows 2003 AD Domain, where users was placed in more than one
5128 Organization Units. When I tried to give exim partial DN, the exit code
5129 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5130 But simultaneously result of request was absolutely normal ldap result,
5131 so I produce this patch..."
5133 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5134 REFERENCE, so I have modified the code to exclude the patch when that macro
5137 55. Some experimental protocols are using DNS PTR records for new purposes. The
5138 keys for these records are domain names, not reversed IP addresses. The
5139 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5140 leaves it alone. Component reversal etc. now happens only for IP addresses.
5143 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5145 57. Double the size of the debug message buffer (to 2048) so that more of very
5146 long debug lines gets shown.
5148 58. The exicyclog utility now does better if the number of log files to keep
5149 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5151 59. Two changes related to the smtp_active_hostname option:
5153 (1) $smtp_active_hostname is now available as a variable.
5154 (2) The default for smtp_banner uses $smtp_active_hostname instead
5155 of $primary_hostname.
5157 60. The host_aton() function is supposed to be passed a string that is known
5158 to be a valid IP address. However, in the case of IPv6 addresses, it was
5159 not checking this. This is a hostage to fortune. Exim now panics and dies
5160 if the condition is not met. A case was found where this could be provoked
5161 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5162 components; fortuitously, this particular loophole had already been fixed
5163 by change 4.50/55 above.
5165 If there are any other similar loopholes, the new check in host_aton()
5166 itself should stop them being exploited. The report I received stated that
5167 data on the command line could provoke the exploit when Exim was running as
5168 exim, but did not say which command line option was involved. All I could
5169 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5170 running as the user.
5173 61. There was a buffer overflow vulnerability in the SPA authentication code
5174 (which came originally from the Samba project). I have added a test to the
5175 spa_base64_to_bits() function which I hope fixes it.
5178 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5179 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5181 63. The daemon start-up calls getloadavg() while still root for those OS that
5182 need the first call to be done as root, but it missed one case: when
5183 deliver_queue_load_max is set with deliver_drop_privilege. This is
5184 necessary for the benefit of the queue runner, because there is no re-exec
5185 when deliver_drop_privilege is set.
5187 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5188 This has been fixed.
5190 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5191 was in use, was not putting the data itself into the right store pool;
5192 consequently, it could be overwritten for a subsequent message in the same
5193 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5196 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5198 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5199 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5200 the function were treating the return as a boolean value, which happened to
5201 work because 0=false and not-0=true, but is not correct code.
5203 68. The host_aton() function was not handling scoped IPv6 addresses (those
5204 with, for example, "%eth0" on the end) correctly.
5206 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5207 negated items (that is, ~something) in unsigned ints. Some compilers
5208 apparently mutter when there is no cast.
5210 70. If an address verification called from an ACL failed, and did not produce a
5211 user-specific message (i.e. there was only a "system" message), nothing was
5212 put in $acl_verify_message. In this situation, it now puts the system
5215 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5216 session; change 4.31/43 added the unwanted input to the log line - except
5217 that it did not do this in the start of session case. It now does.
5219 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5220 This is wrong and can cause the other end to generate a synchronization
5221 error if it is another Exim or anything else that does the synchronization
5222 check. A QUIT command is no longer sent after a timeout.
5224 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5225 during host lookups.
5227 74. The defer_ok option of callout verification was not working if it was used
5228 when verifying addresses in header lines, that is, for this case:
5230 verify = header_sender/callout=defer_ok
5232 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5233 those file descriptors could be used for SMTP connections. If anything
5234 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5235 could be sent to the SMTP client, causing chaos. The daemon now opens
5236 stdin, stdout, and stderr to /dev/null when it puts itself into the
5239 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5240 log by default. The user can override this with suitable Perl magic.
5242 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5243 the log message when discard triggers, was not working for the DATA ACL or
5244 for the non-SMTP ACL.
5246 78. Error message wording change in sieve.c.
5248 79. If smtp_accept_max_per_host was set, the number of connections could be
5249 restricted to fewer than expected, because the daemon was trying to set up
5250 a new connection before checking whether the processes handling previous
5251 connections had finished. The check for completed processes is now done
5252 earlier. On busy systems, this bug wouldn't be noticed because something
5253 else would have woken the daemon, and it would have reaped the completed
5256 80. If a message was submitted locally by a user whose login name contained one
5257 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5258 It caused a spool format error. I have fixed the spool reading code. A
5259 related problem was that the "from" clause in the Received: line became
5260 illegal because of the space(s). It is now covered by ${quote_local_part.
5262 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5265 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5267 83. Patch from Sieve maintainer to fix unterminated string problem in
5268 "vacation" handling.
5270 84. Some minor changes to the Linux configuration files to help with other
5271 OS variants using glibc.
5273 85. One more patch for Sieve to update vacation handling to latest spec.
5276 ----------------------------------------------------
5277 See the note above about the 4.44 and 4.50 releases.
5278 ----------------------------------------------------
5284 1. Change 4.43/35 introduced a bug that caused file counts to be
5285 incorrectly computed when quota_filecount was set in an appendfile
5288 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5289 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5292 3. Give more explanation in the error message when the command for a transport
5293 filter fails to execute.
5295 4. There are several places where Exim runs a non-Exim command in a
5296 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5297 was being done only for the command run by the queryprogram router. It is
5298 now done for all such subprocesses. The other cases are: ${run, transport
5299 filters, and the commands run by the lmtp and pipe transports.
5301 5. Some older OS have a limit of 256 on the maximum number of file
5302 descriptors. Exim was using setrlimit() to set 1000 as a large value
5303 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5304 systems. I've change it so that if it can't get 1000, it tries for 256.
5306 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5307 was an oversight, and furthermore, ever since the addition of extra
5308 controls (e.g. 4.43/32), the checks on when to allow different forms of
5309 "control" were broken. There should now be diagnostics for all cases when a
5310 control that does not make sense is encountered.
5312 7. $recipients is now available in the predata ACL (oversight).
5314 8. Tidy the search cache before the fork to do a delivery from a message
5315 received from the command line. Otherwise the child will trigger a lookup
5316 failure and thereby defer the delivery if it tries to use (for example) a
5317 cached ldap connection that the parent has called unbind on.
5319 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5320 of $address_data from the recipient verification was clobbered by the
5321 sender verification.
5323 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5324 was its contents. (It was OK if the option was not defined at all.)
5326 11. A "Completed" log line is now written for messages that are removed from
5327 the spool by the -Mrm option.
5329 12. $host_address is now set to the target address during the checking of
5330 ignore_target_hosts.
5332 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5333 being passed; this would have caused $sender_host_name to have been used if
5334 matching the list had actually called for a host name (not very likely,
5335 since this list is usually IP addresses). A host name is now passed as
5338 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5339 code that specifies a non-restarting handler (typically sigaction() in
5340 modern systems) in an attempt to fix a rare and obscure crash bug.
5342 15. Narrowed the window for a race in the daemon that could cause it to ignore
5343 SIGCHLD signals. This is not a major problem, because they are used only to
5344 wake it up if nothing else does.
5346 16. A malformed maildirsize file could cause Exim to calculate negative values
5347 for the mailbox size or file count. Odd effects could occur as a result.
5348 The maildirsize information is now recalculated if the size or filecount
5351 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5352 support for a long time. Removed HAVE_SYS_VFS_H.
5354 18. Updated exipick to current release from John Jetmore.
5356 19. Allow an empty sender to be matched against a lookup in an address list.
5357 Previously the only cases considered were a regular expression, or an
5360 20. Exim went into a mad DNS lookup loop when doing a callout where the
5361 host was specified on the transport, if the DNS lookup yielded more than
5364 21. The RFC2047 encoding function was originally intended for short strings
5365 such as real names; it was not keeping to the 75-character limit for
5366 encoded words that the RFC imposes. It now respects the limit, and
5367 generates multiple encoded words if necessary. To be on the safe side, I
5368 have increased the buffer size for the ${rfc2047: expansion operator from
5371 22. Failure to deliver a bounce message always caused it to be frozen, even if
5372 there was an errors_to setting on the router. The errors_to setting is now
5375 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5376 canonical form (fully expanded) before being placed in
5377 $sender_host_address.
5379 24. Updated eximstats to version 1.33
5381 25. Include certificate and key file names in error message when GnuTLS fails
5382 to set them up, because the GnuTLS error message doesn't include the name
5383 of the failing file when there is a problem reading it.
5385 26. Expand error message when OpenSSL has problems setting up cert/key files.
5388 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5389 (this can affect the format of dates).
5391 28. exim_tidydb, when checking for the continued existence of a message for
5392 which it has found a message-specific retry record, was not finding
5393 messages that were in split spool directories. Consequently, it was
5394 deleting retry records that should have stayed in existence.
5396 29. eximstats updated to version 1.35
5397 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5398 1.35 - bugfix such that pie charts by volume are generated correctly
5400 30. The SPA authentication driver was not abandoning authentication and moving
5401 on to the next authenticator when an expansion was forced to fail,
5402 contradicting the general specification for all authenticators. Instead it
5403 was generating a temporary error. It now behaves as specified.
5405 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5406 (the order specifies the preference for clients). The order is now AES256,
5407 AES128, 3DES, ARCFOUR128.
5409 31. Small patch to Sieve code - explicitly set From: when generating an
5412 32. Exim crashed if a remote delivery caused a very long error message to be
5413 recorded - for instance if somebody sent an entire SpamAssassin report back
5414 as a large number of 550 error lines. This bug was coincidentally fixed by
5415 increasing the size of one of Exim's internal buffers (big_buffer) that
5416 happened as part of the Exiscan merge. However, to be on the safe side, I
5417 have made the code more robust (and fixed the comments that describe what
5420 33. Some experimental protocols are using DNS PTR records for new purposes. The
5421 keys for these records are domain names, not reversed IP addresses. The
5422 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5423 leaves it alone. Component reversal etc. now happens only for IP addresses.
5426 34. The host_aton() function is supposed to be passed a string that is known
5427 to be a valid IP address. However, in the case of IPv6 addresses, it was
5428 not checking this. This is a hostage to fortune. Exim now panics and dies
5429 if the condition is not met. A case was found where this could be provoked
5430 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5431 components; fortuitously, this particular loophole had already been fixed
5432 by change 4.50/55 or 4.44/33 above.
5434 If there are any other similar loopholes, the new check in host_aton()
5435 itself should stop them being exploited. The report I received stated that
5436 data on the command line could provoke the exploit when Exim was running as
5437 exim, but did not say which command line option was involved. All I could
5438 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5439 running as the user.
5442 35. There was a buffer overflow vulnerability in the SPA authentication code
5443 (which came originally from the Samba project). I have added a test to the
5444 spa_base64_to_bits() function which I hope fixes it.
5447 36. The daemon start-up calls getloadavg() while still root for those OS that
5448 need the first call to be done as root, but it missed one case: when
5449 deliver_queue_load_max is set with deliver_drop_privilege. This is
5450 necessary for the benefit of the queue runner, because there is no re-exec
5451 when deliver_drop_privilege is set.
5453 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5454 was in use, was not putting the data itself into the right store pool;
5455 consequently, it could be overwritten for a subsequent message in the same
5456 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5459 38. Sometimes the final signoff response after QUIT could fail to get
5460 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5461 < 0 before doing a fflush(). This bug looks as though it goes back to the
5462 introduction of TLS in release 3.20, but "sometimes" must have been rare
5463 because the tests only now provoked it.
5469 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5470 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5471 Now it returns an integer. A number of calls were still expecting a T/F
5472 return. Fortuitously, in all cases, the tests worked in OK situations,
5473 which is the norm. However, things would have gone wrong on any write
5474 failures on the smtp file descriptor. This function is used when sending
5475 messages over SMTP and also when doing verify callouts.
5477 2. When Exim is called to do synchronous delivery of a locally submitted
5478 message (the -odf or -odi options), it no longer closes stderr before doing
5481 3. Implemented the mua_wrapper option.
5483 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5485 5. Implemented the functions header_remove(), header_testname(),
5486 header_add_at_position(), and receive_remove_recipient(), and exported them
5489 6. If an ACL "warn" statement specified the addition of headers, Exim already
5490 inserted X-ACL-Warn: at the start if there was no header name. However, it
5491 was not making this test for the second and subsequent header lines if
5492 there were newlines in the string. This meant that an invalid header could
5493 be inserted if Exim was badly configured.
5495 7. Allow an ACL "warn" statement to add header lines at the start or after all
5496 the Received: headers, as well as at the end.
5498 8. Added the rcpt_4xx retry error code.
5500 9. Added postmaster_mailfrom=xxx to callout verification option.
5502 10. Added mailfrom=xxxx to the callout verification option, for verify=
5505 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5506 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5508 12. Inserted some casts to stop certain compilers warning when using pointer
5509 differences as field lengths or precisions in printf-type calls (mostly
5510 affecting debugging statements).
5512 13. Added optional readline() support for -be (dynamically loaded).
5514 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5515 same clock tick as a message's arrival, so that its received time was the
5516 same as the "first fail" time on the retry record, and that message
5517 remained on the queue past the ultimate address timeout, every queue runner
5518 would try a delivery (because it was past the ultimate address timeout) but
5519 after another failure, the ultimate address timeout, which should have then
5520 bounced the address, did not kick in. This was a "< instead of <=" error;
5521 in most cases the first failure would have been in the next clock tick
5522 after the received time, and all would be well.
5524 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5525 being recognized when the domain list was tested by the match_domain
5526 condition in an expansion string.
5528 16. Added the ${str2b64: operator.
5530 17. Exim was always calling setrlimit() to set a large limit for the number of
5531 processes, without checking whether the existing limit was already
5532 adequate. (It did check for the limit on file descriptors.) Furthermore,
5533 errors from getrlimit() and setrlimit() were being ignored. Now they are
5534 logged to the main and panic logs, but Exim does carry on, to try to do its
5535 job under whatever limits there are.
5537 18. Imported PCRE 5.0.
5539 19. Trivial typo in log message " temporarily refused connection" (the leading
5542 20. If the log selector return_path_on_delivery was set and an address was
5543 redirected to /dev/null, the delivery process crashed because it assumed
5544 that a return path would always be set for a "successful" delivery. In this
5545 case, the whole delivery is bypassed as an optimization, and therefore no
5548 21. Internal re-arrangement: the function for sending a challenge and reading
5549 a response while authentication was assuming a zero-terminated challenge
5550 string. It's now changed to take a pointer and a length, to allow for
5551 binary data in such strings.
5553 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5555 23. Exim was not respecting finduser_retries when seeking the login of the
5556 uid under which it was called; it was always trying 10 times. (The default
5557 setting of finduser_retries is zero.) Also, it was sleeping after the final
5558 failure, which is pointless.
5560 24. Implemented tls_on_connect_ports.
5562 25. Implemented acl_smtp_predata.
5564 26. If the domain in control=submission is set empty, Exim assumes that the
5565 authenticated id is a complete email address when it generates From: or
5566 Sender: header lines.
5568 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5569 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5570 chown and chgrp in /bin and hostname in /usr/bin.
5572 28. Exim was keeping the "process log" file open after each use, just as it
5573 does for the main log. This opens the possibility of it remaining open for
5574 long periods when the USR1 signal hits a daemon. Occasional processlog
5575 errors were reported, that could have been caused by this. Anyway, it seems
5576 much more sensible not to leave this file open at all, so that is what now
5579 29. The long-running daemon process does not normally write to the log once it
5580 has entered its main loop, and it closes the log before doing so. This is
5581 so that log files can straightforwardly be renamed and moved. However,
5582 there are a couple of unusual error situations where the daemon does write
5583 log entries, and I had neglected to close the log afterwards.
5585 30. The text of an SMTP error response that was received during a remote
5586 delivery was being truncated at 512 bytes. This is too short for some of
5587 the long messages that one sometimes sees. I've increased the limit to
5590 31. It is now possible to make retry rules that apply only when a message has a
5591 specific sender, in particular, an empty sender.
5593 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5594 it possible to be selective about when SMTP synchronization is enforced.
5596 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5598 32. Implemented hosts_connection_nolog.
5600 33. Added an ACL for QUIT.
5602 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5605 35. Added mailbox_size and mailbox_filecount to appendfile.
5607 36. Added control = no_multiline_responses to ACLs.
5609 37. There was a bug in the logic of the code that waits for the clock to tick
5610 in the case where the clock went backwards by a substantial amount such
5611 that the microsecond fraction of "now" was more than the microsecond
5612 fraction of "then" (but the whole seconds number was less).
5614 38. Added support for the libradius Radius client library this is found on
5615 FreeBSD (previously only the radiusclient library was supported).
5621 1. When certain lookups returned multiple values in the form name=value, the
5622 quoting of the values was not always being done properly. Specifically:
5623 (a) If the value started with a double quote, but contained no whitespace,
5625 (b) If the value contained whitespace other than a space character (i.e.
5626 tabs or newlines or carriage returns) it was not quoted.
5627 This fix has been applied to the mysql and pgsql lookups by writing a
5628 separate quoting function and calling it from the lookup code. The fix
5629 should probably also be applied to nisplus, ibase and oracle lookups, but
5630 since I cannot test any of those, I have not disturbed their existing code.
5632 2. A hit in the callout cache for a specific address caused a log line with no
5633 reason for rejecting RCPT. Now it says "Previous (cached) callout
5634 verification failure".
5636 3. There was an off-by-one bug in the queryprogram router. An over-long
5637 return line was truncated at 256 instead of 255 characters, thereby
5638 overflowing its buffer with the terminating zero. As well as fixing this, I
5639 have increased the buffer size to 1024 (and made a note to document this).
5641 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5642 when Exim is waiting for an SMTP response from a remote server, Exim
5643 restarts its select() call on the socket, thereby resetting its timeout.
5644 This is not a problem when such interrupts are rare. Somebody set up a cron
5645 job to run exiwhat every 2 minutes, which is less than the normal select()
5646 timeout (5 or 10 minutes). This meant that the select() timeout never
5647 kicked in because it was always reset. I have fixed this by comparing the
5648 time when an interrupt arrives with the time at the start of the first call
5649 to select(). If more time than the timeout has elapsed, the interrupt is
5650 treated as a timeout.
5652 5. Some internal re-factoring in preparation for the addition of Sieve
5653 extensions (by MH). In particular, the "personal" test is moved to a
5654 separate function, and given an option for scanning Cc: and Bcc: (which is
5655 not set for Exim filters).
5657 6. When Exim created an email address using the login of the caller as the
5658 local part (e.g. when creating a From: or Sender: header line), it was not
5659 quoting the local part when it contained special characters such as @.
5661 7. Installed new OpenBSD configuration files.
5663 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5664 try to make them clearer.
5666 9. Callout options, other than the timeout value, were being ignored when
5667 verifying sender addresses in header lines. For example, when using
5669 verify = header_sender/callout=no_cache
5671 the cache was (incorrectly) being used.
5673 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5674 only the Cygwin environment.
5676 11. When return_path_on_delivery was set as a log selector, if different remote
5677 addresses in the same message used different return paths and parallel
5678 remote delivery occurred, the wrong values would sometimes be logged.
5679 (Whenever a remote delivery process finished, the return path value from
5680 the most recently started remote delivery process was logged.)
5682 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5683 lines when AUTH and/or TLS are in use. This is the "received protocol"
5684 field. Exim used to use "asmtp" for authenticated SMTP, without any
5685 indication (in the protocol name) for TLS use. Now it follows the RFC and
5686 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5687 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5688 names appear in log lines as well as in Received: header lines.
5690 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5691 extensions, and comparison tests, and to fix some bugs.
5693 14. Changes to the "personal" filter test:
5695 (1) The test was buggy in that it was just doing the equivalent of
5696 "contains" tests on header lines. For example, if a user's address was
5697 anne@some.where, the "personal" test would incorrectly be true for
5699 To: susanne@some.where
5701 This test is now done by extracting each address from the header in turn,
5702 and checking the entire address. Other tests that are part of "personal"
5703 are now done using regular expressions (for example, to check local parts
5704 of addresses in From: header lines).
5706 (2) The list of non-personal local parts in From: addresses has been
5707 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5708 taken from the Sieve specification recommendations.
5710 (3) If the message contains any header line starting with "List-" it is
5711 treated as non-personal.
5713 (4) The test for "circular" in the Subject: header line has been removed
5714 because it now seems ill-conceived.
5716 15. Minor typos in src/EDITME comments corrected.
5718 16. Installed latest exipick from John Jetmore.
5720 17. If headers_add on a router specified a text string that was too long for
5721 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5722 of string_sprintf() is now avoided.
5724 18. $message_body_size was not set (it was always zero) when running the DATA
5725 ACL and the local_scan() function.
5727 19. For the "mail" command in an Exim filter, no default was being set for
5728 the once_repeat time, causing a random time value to be used if "once" was
5729 specified. (If the value happened to be <= 0, no repeat happened.) The
5730 default is now 0s, meaning "never repeat". The "vacation" command was OK
5731 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5732 (I found it when inspecting the code).
5734 20. There is now an overall timeout for performing a callout verification. It
5735 defaults to 4 times the callout timeout, which applies to individual SMTP
5736 commands during the callout. The overall timeout applies when there is more
5737 than one host that can be tried. The timeout is checked before trying the
5738 next host. This prevents very long delays if there are a large number of
5739 hosts and all are timing out (e.g. when the network connections are timing
5740 out). The value of the overall timeout can be changed by specifying an
5741 additional sub-option for "callout", called "maxwait". For example:
5743 verify = sender/callout=5s,maxwait=20s
5745 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5746 to the end before writing, but this should make it even safer).
5748 22. Exim was forgetting that it had advertised PIPELINING for the second and
5749 subsequent messages on an SMTP connection. It was also not resetting its
5750 memory on STARTTLS and an internal HELO.
5752 23. When Exim logs an SMTP synchronization error within a session, it now
5753 records whether PIPELINING has been advertised or not.
5755 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5756 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5757 rather than long int.
5759 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5761 26. Added the never_mail option to autoreply.
5767 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5768 crash if the getsockname() call failed; this can happen if a connection is
5769 closed very soon after it is established. The problem was simply in the
5770 order in which certain operations were done, causing Exim to try to write
5771 to the SMTP stream before it had set up the file descriptor. The bug has
5772 been fixed by making things happen in the correct order.
5778 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5779 before the connection was closed, thus losing the rejection response.
5781 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5782 some early Solaris releases, but causes trouble in current releases where
5783 socklen_t is defined.
5785 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5788 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5791 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5792 that releasing the top of it at the end releases what was used for sub-
5793 expansions (unless the block got too big). However, discard this block if
5794 the first thing is a variable or header, so that we can use its block when
5795 it is dynamic (useful for very large $message_headers, for example).
5797 6. Lookups now cache *every* query, not just the most recent. A new, separate
5798 store pool is used for this. It can be recovered when all lookup caches are
5799 flushed. Lookups now release memory at the end of their result strings.
5800 This has involved some general refactoring of the lookup sources.
5802 7. Some code has been added to the store_xxx() functions to reduce the amount
5803 of flapping under certain conditions.
5805 8. log_incoming_interface used to affect only the <= reception log lines. Now
5806 it causes the local interface and port to be added to several more SMTP log
5807 lines, for example "SMTP connection from", and rejection lines.
5809 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5811 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5813 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5814 could be overwritten at the end of the current message (or the start of a
5815 new message if it was set in a HELO ACL). The value is now preserved for
5816 the duration of the SMTP connection.
5818 12. If a transport had a headers_rewrite setting, and a matching header line
5819 contained an unqualified address, that address was qualified, even if it
5820 did not match any rewriting rules. The underlying bug was that the values
5821 of the flags that permit the existence of unqualified sender and recipient
5822 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5823 non-local messages, and by -bnq for local messages) were not being
5824 preserved with the message after it was received.
5826 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5827 "next input=" as part of the text comprising the host identity instead of
5828 the correct text. The code was using the same buffer for two different
5829 strings. However, depending on which order the printing function evaluated
5830 its arguments, the bug did not always show up. Under Linux, for example, my
5831 test suite worked just fine.
5833 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5834 doesn't work with some older versions of Perl. It has been changed to "my",
5835 which in any case is probably the better facility to use.
5837 15. A really picky compiler found some instances of statements for creating
5838 error messages that either had too many or two few arguments for the format
5841 16. The size of the buffer for calls to the DNS resolver has been increased
5842 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5843 for addresses that have a lot of PTR records. This alleviates a problem; it
5844 does not fully solve it.
5846 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5847 buffer now truncates the list and logs the incident, which is the same
5848 action as happens when Exim is looking up a host name and its aliases.
5849 Previously in this situation something unpredictable would happen;
5850 sometimes it was "internal error: store_reset failed".
5852 18. If a server dropped the connection unexpectedly when an Exim client was
5853 using GnuTLS and trying to read a response, the client delivery process
5854 crashed while trying to generate an error log message.
5856 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5857 string, for example:
5859 warn message = H1: something\nH2: something
5861 the text was added as a single header line from Exim's point of view
5862 though it ended up OK in the delivered message. However, searching for the
5863 second and subsequent header lines using $h_h2: did not work. This has been
5864 fixed. Similarly, if a system filter added multiple headers in this way,
5865 the routers could not see them.
5867 20. Expanded the error message when iplsearch is called with an invalid key to
5868 suggest using net-iplsearch in a host list.
5870 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5871 ACLs are no longer actually imposed (and a message to that effect is
5874 22. If a "gecos" field in a passwd entry contained escaped characters, in
5875 particular, if it contained a \" sequence, Exim got it wrong when building
5876 a From: or a Sender: header from that name. A second bug also caused
5877 incorrect handling when an unquoted " was present following a character
5878 that needed quoting.
5880 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5881 was not being matched caselessly.
5883 24. Arranged for all hyphens in the exim.8 source to be escaped with
5886 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5887 properly. Recipient callout cache records were still being keyed to include
5888 the sender, even when use_sender was set false. This led to far more
5889 callouts that were necessary. The sender is no longer included in the key
5890 when use_sender is false.
5892 26. Added "control = submission" modifier to ACLs.
5894 27. Added the ${base62d: operator to decode base 62 numbers.
5896 28. dnsdb lookups can now access SRV records.
5898 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5899 the configuration file.
5901 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5902 (-v) mode. This makes the output for a verbose queue run more intelligible.
5904 31. Added a use_postmaster feature to recipient callouts.
5906 32. Added the $body_zerocount variable, containing the number of binary zero
5907 bytes in the message body.
5909 33. The time of last modification of the "new" subdirectory is now used as the
5910 "mailbox time last read" when there is a quota error for a maildir
5913 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5915 35. Added +ignore_unknown as a special item in host lists.
5917 36. Code for decoding IPv6 addresses in host lists is now included, even if
5918 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5919 address was recognized as an IP address, but was then not correctly decoded
5920 into binary, causing unexpected and incorrect effects when compared with
5927 1. Very minor rewording of debugging text in manualroute to say "list of
5928 hosts" instead of "hostlist".
5930 2. If verify=header_syntax was set, and a header line with an unqualified
5931 address (no domain) and a large number of spaces between the end of the
5932 name and the colon was received, the reception process suffered a buffer
5933 overflow, and (when I tested it) crashed. This was caused by some obsolete
5934 code that should have been removed. The fix is to remove it!
5936 3. When running in the test harness, delay a bit after writing a bounce
5937 message to get a bit more predictability in the log output.
5939 4. Added a call to search_tidyup() just before forking a reception process. In
5940 theory, someone could use a lookup in the expansion of smtp_accept_max_
5941 per_host which, without the tidyup, could leave open a database connection.
5943 5. Added the variables $recipient_data and $sender_data which get set from a
5944 lookup success in an ACL "recipients" or "senders" condition, or a router
5945 "senders" option, similar to $domain_data and $local_part_data.
5947 6. Moved the writing of debug_print from before to after the "senders" test
5950 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5951 problems for message scanning, either using a data ACL, or using
5952 local_scan() because the Received: header was not generated till after they
5953 were called (in order to set the time as the time of reception completion).
5954 I have revised the way this works. The header is now generated after the
5955 body is received, but before the ACL or local_scan() are called. After they
5956 are run, the timestamp in the header is updated.
5962 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5963 before starting a queue runner without re-exec. This happened only when
5964 deliver_drop_privilege was set or when the Exim user was set to root. The
5965 effect of the bug was that timeouts during subsequent deliveries caused
5966 crashes instead of being properly handled. The handler is now left at its
5967 default (and expected) setting.
5969 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5970 message, again when deliver_drop_privilege is set or Exim is run as root.
5971 The bug described in (1) was not present in this case, but the tidying up
5972 of the other signals was missing. I have made the two cases consistent.
5974 3. The ignore_target_hosts setting on a manualroute router was being ignored
5975 for hosts that were looked up using the /MX notation.
5977 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5980 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5981 operated on the sender address. After changing the $sender_address to <>
5982 for the sender address verify, Exim was re-instated it as the original
5983 (before rewriting) address, but remembering that it had rewritten it, so it
5984 wasn't rewriting it again. This bug also had the effect of breaking the
5985 sender address verification caching when the sender address was rewritten.
5987 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5988 This has been changed so that if the ip literal address matches
5989 ignore_target_hosts, the router declines.
5991 7. Added expansion conditions match_domain, match_address, and match_local_
5992 part (NOT match_host).
5994 8. The placeholder for the Received: header didn't have a length field set.
5996 9. Added code to Exim itself and to exim_lock to test for a specific race
5997 condition that could lead to file corruption when using MBX delivery. The
5998 issue is with the lockfile that is created in /tmp. If this file is removed
5999 after a process has opened it but before that process has acquired a lock,
6000 there is the potential for a second process to recreate the file and also
6001 acquire a lock. This could lead to two Exim processes writing to the file
6002 at the same time. The added code performs the same test as UW imapd; it
6003 checks after acquiring the lock that its file descriptor still refers to
6004 the same named file.
6006 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6007 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6010 11. Added the smtp_active_hostname option. If used, this will typically be made
6011 to depend on the incoming interface address. Because $interface_address is
6012 not set up until the daemon has forked a reception process, error responses
6013 that can happen earlier (such as "too many connections") no longer contain
6016 12. If an expansion in a condition on a "warn" statement fails because a lookup
6017 defers, the "warn" statement is abandoned, and the next ACL statement is
6018 processed. Previously this caused the whole ACL to be aborted.
6020 13. Added the iplsearch lookup type.
6022 14. Added ident_timeout as a log selector.
6024 15. Added tls_certificate_verified as a log selector.
6026 16. Added a global option tls_require_ciphers (compare the smtp transport
6027 option of the same name). This controls incoming TLS connections.
6029 17. I finally figured out how to make tls_require_ciphers do a similar thing
6030 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6031 before starting the TLS session.
6033 18. Tabs are now shown as \t in -bP output.
6035 19. If the log selector return_path_on_delivery was set, Exim crashed when
6036 bouncing a message because it had too many Received: header lines.
6038 20. If two routers both had headers_remove settings, and the first one included
6039 a superfluous trailing colon, the final name in the first list and the
6040 first name in the second list were incorrectly joined into one item (with a
6041 colon in the middle).
6047 1. Added -C and -D options to the exinext utility, mainly to make it easier
6048 to include in the automated testing, but these could be helpful when
6049 multiple configurations are in use.
6051 2. The exinext utility was not formatting the output nicely when there was
6052 an alternate port involved in the retry record key, nor when there was a
6053 message id as well (for retries that were specific to a specific message
6054 and a specific host). It was also confused by IPv6 addresses, because of
6055 the additional colons they contain. I have fixed the IPv4 problem, and
6056 patched it up to do a reasonable job for IPv6.
6058 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6059 delivery, the log line now contains "pipelined" if PIPELINING was used.
6061 4. An SMTP transport process used to panic and die if the bind() call to set
6062 an explicit outgoing interface failed. This has been changed; it is now
6063 treated in the same way as a connect() failure.
6065 5. A reference to $sender_host_name in the part of a conditional expansion
6066 that was being skipped was still causing a DNS lookup. This no longer
6069 6. The def: expansion condition was not recognizing references to header lines
6070 that used bh_ and bheader_.
6072 7. Added the _cache feature to named lists.
6074 8. The code for checking quota_filecount in the appendfile transport was
6075 allowing one more file than it should have been.
6077 9. For compatibility with Sendmail, the command line option
6085 and sets the incoming protocol and host name (for trusted callers). The
6086 host name and its colon can be omitted when only the protocol is to be set.
6087 Note the Exim already has two private options, -pd and -ps, that refer to
6088 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6089 "s", but I don't think that's a major issue.
6091 10. A number of refactoring changes to the code, none of which should affect
6094 (a) The number of logging options was getting close to filling up the
6095 32-bit word that was used as a bit map. I have split them into two classes:
6096 those that are passed in the argument to log_write(), and those that are
6097 only ever tested independently outside of that function. These are now in
6098 separate 32-bit words, so there is plenty of room for expansion again.
6099 There is no change in the user interface or the logging behaviour.
6101 (b) When building, for example, log lines, the code previously used a
6102 macro that called string_cat() twice, in order to add two strings. This is
6103 not really sufficiently general. Furthermore, there was one instance where
6104 it was actually wrong because one of the argument was used twice, and in
6105 one call a function was used. (As it happened, calling the function twice
6106 did not affect the overall behaviour.) The macro has been replaced by a
6107 function that can join an arbitrary number of extra strings onto a growing
6110 (c) The code for expansion conditions now uses a table and a binary chop
6111 instead of a serial search (which was left over from when there were very
6112 few conditions). Also, it now recognizes conditions like "pam" even when
6113 the relevant support is not compiled in: a suitably worded error message is
6114 given if an attempt is made to use such a condition.
6116 11. Added ${time_interval:xxxxx}.
6118 12. A bug was causing one of the ddress fields not to be passed back correctly
6119 from remote delivery subprocesses. The field in question was not being
6120 subsequently used, so this caused to problems in practice.
6122 13. Added new log selectors queue_time and deliver_time.
6124 14. Might have fixed a bug in maildirsizefile handling that threw up
6125 "unexpected character" debug warnings, and recalculated the data
6126 unnecessarily. In any case, I expanded the warning message to give more
6129 15. Added the message "Restricted characters in address" to the statements in
6130 the default ACL that block characters like @ and % in local parts.
6132 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6133 Three changes have been made:
6135 (a) There was a serious bug; a negative response to MAIL caused the whole
6136 recipient domain to be cached as invalid, thereby blocking all messages
6137 to all local parts at the same domain, from all senders. This bug has
6138 been fixed. The domain is no longer cached after a negative response to
6139 MAIL if the sender used is not empty.
6141 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6144 (c) A new callout option, "use_sender" has been added for people who want
6145 the modified behaviour.
6151 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6154 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6155 indeed breaks things for older releases.
6157 3. Added additional logging to the case where there is a problem reading data
6158 from a filter that is running in a subprocess using a pipe, in order to
6159 try to track down a specific problem.
6161 4. Testing facility fudge: when running in the test harness and attempting
6162 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6163 getting "No route to host". Convert this to a timeout.
6165 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6168 6. Some OS don't have socklen_t but use size_t instead. This affects the
6169 fifth argument of getsockopt() amongst other things. This is now
6170 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6171 can be set for individual OS. I have set it for SunOS5, OSF1, and
6172 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6173 some earlier ones do not.
6175 7. Change 4.30/15 was not doing the test caselessly.
6177 8. The standard form for an IPv6 address literal was being rejected by address
6178 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6179 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6180 this, as well as the form without the "IPv6" on the front (but only when
6181 address literals are enabled, of course).
6183 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6185 10. Exim crashed if a message with an empty sender address specified by -f
6186 encountered a router with an errors_to setting. This could be provoked only
6187 by a command such as
6191 where an empty string was supplied; "<>" did not hit this bug.
6193 11. Installed PCRE release 4.5.
6195 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6196 remained set. It is now erased.
6198 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6199 times from message ids (which are base 36 rather than the normal 62).
6201 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6202 were being counted as actual protocol errors, and logged if the log
6203 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6204 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6205 and DATA following a set of rejected RCPTs do not count as protocol errors.
6206 In other words, Exim assumes they were pipelined, though this may not
6207 actually be the case. Of course, in all cases the client gets an
6208 appropriate error code.
6210 15. If a lookup fails in an ACL condition, a message about the failure may
6211 be available; it is used if testing the ACL cannot continue, because most
6212 such messages specify what the cause of the deferral is. However, some
6213 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6214 that caused an old message to be retained and used if a later statement
6215 caused a defer, replacing the real cause of the deferral.
6217 16. If an IP address had so many PTR records that the DNS lookup buffer
6218 was not large enough to hold them, Exim could crash while trying to process
6219 the truncated data. It now detects and logs this case.
6221 17. Further to 4.21/58, another change has been made: if (and only if) the
6222 first line of a message (the first header line) ends with CRLF, a bare LF
6223 in a subsequent header line has a space inserted after it, so as not to
6224 terminate the header.
6226 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6227 unnecessarily, used atoi() instead of strtol(), and didn't check the
6228 termination when getting file sizes from file names by regex.
6230 19. Completely re-implemented the support for maildirsize files, in the light
6231 of a number of problems with the previous contributed implementation
6232 (4.30/29). In particular:
6234 . If the quota is zero, the maildirsize file is maintained, but no quota is
6237 . If the maildir directory does not exist, it is created before any attempt
6238 to write a maildirsize file.
6240 . The quota value in the file is just a cache; if the quota is changed in
6241 the transport, the new value overrides.
6243 . A regular expression is available for excluding directories from the
6246 20. The autoreply transport checks the characters in options that define the
6247 message's headers; it allows continued headers, but it was checking with
6248 isspace() after an embedded newline instead of explicitly looking for a
6251 21. If all the "regular" hosts to which an address was routed had passed their
6252 expiry times, and had not reached their retry times, the address was
6253 bounced, even if fallback hosts were defined. Now Exim should go on to try
6256 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6257 equivalent code in the SMTP transport. Some hosts send humungous responses
6258 to HELO/EHLO, more than 1024 it seems.
6260 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6261 gives compiler warnings in some environments. I've now done it "properly",
6264 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6265 (because of problems with the built-in one) was declared to return uschar *
6266 instead of char *, causing compiler failure.
6268 25. Fixed a file descriptor leak when processing alias/forward files.
6270 26. Fixed a minor format string issue in dbfn.c.
6272 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6274 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6275 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6278 29. When a sender address is verified, it is cached, to save repeating the test
6279 when there is more than one recipient in a message. However, when the
6280 verification involves a callout, it is possible for different callout
6281 options to be set for different recipients. It is too complicated to keep
6282 track of this in the cache, so now Exim always runs a verification when a
6283 callout is required, relying on the callout cache for the optimization.
6284 The overhead is duplication of the address routing, but this should not be
6287 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6288 to be verified with callout=postmaster, and the main callout worked but the
6289 postmaster check failed, the verification correctly failed. However, if a
6290 subsequent RCPT command asked for sender verification *without* the
6291 postmaster check, incorrect caching caused this verification also to fail,
6294 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6295 it was not caching the DNS options (qualify_single, search_parents) that
6296 were used when the lookup failed. A subsequent lookup with different
6297 options therefore always gave the same answer, though there were cases
6298 where it should not have. (Example: a "domains = !$mx_any" option on a
6299 dnslookup router: the "domains" option is always processed without any
6300 widening, but the router might have qualify_single set.) Now Exim uses the
6301 cached value only when the same options are set.
6303 32. Added John Jetmore's "exipick" utility to the distribution.
6305 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6306 than a timeout (e.g. a certificate is required, and is not provided), an
6307 Exim server now closes the connection immediately. Previously it waited for
6308 the client to close - but if the client is SSL, it seems that they each
6309 wait for each other, leading to a delay before one of them times out.
6311 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6312 maintained 0.8.x compatibility because I don't think many are using it, and
6313 it is clearly obsolete.
6315 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6318 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6319 client certificate was expired. A simple patch fixes this, though I don't
6320 understand the full logic of why the verify callback is called multiple
6323 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6324 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6325 which causes problems with some clients (such as the Certicom SSL Plus
6326 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6327 disables the coutermeasure allowing Eudora to connect."
6329 38. Exim was not checking that a write() to a log file succeeded. This could
6330 lead to Bad Things if a log got too big, in particular if it hit a file
6331 size limit. Exim now panics and dies if it cannot write to a log file, just
6332 as it does if it cannot open a log file.
6334 39. Modified OS/Makefile-Linux so that it now contains
6336 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6338 The two -D definitions ensure that Exim is compiled with large file
6339 support, which makes it possible to handle log files that are bigger than
6342 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6343 instance) a domain was checked against a named list that involved a lookup,
6344 causing $domain_data to be set, then another domain was checked against the
6345 same list, then the first domain was re-checked, the value of $domain_data
6346 after the final check could be wrong. In particular, if the second check
6347 failed, it could be set empty. This bug probably also applied to
6350 41. The strip_trailing_dot option was not being applied to the address given
6351 with the -f command-line option.
6353 42. The code for reading a message's header from the spool was incrementing
6354 $received_count, but never initializing it. This meant that the value was
6355 incorrect (doubled) while delivering a message in the same process in which
6356 it was received. In the most common configuration of Exim, this never
6357 happens - a fresh exec is done - but it can happen when
6358 deliver_drop_privilege is set.
6360 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6361 it now includes up to 150 characters of the unexpected data in the log
6364 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6365 and building data strings. The size of both of these buffers was 10 000
6366 bytes - far larger than anybody would *ever* want, thought I. Needless to
6367 say, somebody hit the limit. I have increased the maximum line length to
6368 20 000 and the maximum data length of concatenated lines to 100 000. I have
6369 also fixed two bugs, because there was no checking on these buffers. Tsk,
6370 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6371 buffer is too small.
6373 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6374 lsearch lookups. Now it does.
6376 46. When parsing a route_list item in a manualroute router, a fixed-length
6377 buffer was used for the list of hosts. I made this 1024 bytes long,
6378 thinking that nobody would ever have a list of hosts that long. Wrong.
6379 Somebody had a whole pile of complicated expansion conditions, and the
6380 string was silently truncated, leading to an expansion error. It turns out
6381 that it is easier to change to an unlimited length (owing to other changes
6382 that have happened since this code was originally written) than to build
6383 structure for giving a limitation error. The length of the item that
6384 expands into the list of hosts is now unlimited.
6386 47. The lsearch lookup could not handle data where the length of text line was
6387 more than 4095 characters. Such lines were truncated, leading to shortened
6388 data being returned. It should now handle lines of any length.
6390 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6391 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6394 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6395 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6396 no longer happens in comment lines. A list of replacements is now placed
6397 at the head of all of the source files, except those whose only change is
6398 to replace PERL_COMMAND in the very first #! line.
6400 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6401 messages on the queue, with a bottom-up merge sort, using code contributed
6402 by Michael Haardt. This should make operations like -bp somewhat faster on
6403 large queues. It won't affect queue runners, except when queue_run_in_order
6406 51. Installed eximstats 1.31 in the distribution.
6408 52. Added support for SRV lookups to the dnslookup router.
6410 53. If an ACL referred to $message_body or $message_body_end, the value was not
6411 reset for any messages that followed in the same SMTP session.
6413 54. The store-handling optimization for building very long strings was not
6414 differentiating between the different store pools. I don't think this
6415 actually made any difference in practice, but I've tidied it.
6417 55. While running the routers to verify a sender address, $sender_address
6418 was still set to the sender address. This is wrong, because when routing to
6419 send a bounce to the sender, it would be empty. Therefore, I have changed
6420 it so that, while verifying a sender address, $sender_address is set to <>.
6421 (There is no change to what happens when verifying a recipient address.)
6423 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6424 target A or AAAA records (if not already returned) without resetting the
6425 qualify_single or search_parents options of the DNS resolver. These are
6426 inappropriate in this case because the targets of MX and SRV records must
6427 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6428 target that, when qualified, matched something in the local domain. These
6429 two options are now turned off when doing these lookups.
6431 57. It seems that at least some releases of Reiserfs (which does not have the
6432 concept of a fixed number of inodes) returns zero and not -1 for the
6433 number of available inodes. This interacted badly with check_spool_inodes,
6434 which assumed that -1 was the "no such thing" setting. What I have done is
6435 to check that the total number of inodes is greater than zero before doing
6436 the test of how many are available.
6438 58. When a "warn" ACL statement has a log_message modifier, the message is
6439 remembered, and not repeated. This is to avoid a lot of repetition when a
6440 message has many recipients that cause the same warning to be written.
6441 However, Exim was preserving the list of already written lines for an
6442 entire SMTP session, which doesn't seem right. The memory is now reset if a
6443 new message is started.
6445 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6446 debugging output unless log_rewrite was also set.
6448 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6449 of (int)(handle) when we know that handle contains (void *)(-1).
6451 61. The Exim daemon panic-logs an error return when it closes the incoming
6452 connection. However "connection reset by peer" seems to be common, and
6453 isn't really an error worthy of noting specially, so that particular error
6456 62. When Exim is trying to find all the local interfaces, it used to panic and
6457 die if the ioctl to get the interface flags failed. However, it seems that
6458 on at least one OS (Solaris 9) it is possible to have an interface that is
6459 included in the list of interfaces, but for which you get a failure error
6460 for this call. This happens when the interface is not "plumbed" into a
6461 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6462 failure of the "get flags" call assumes that the interface is down.
6464 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6465 makes life easier for people who are doing arithmetic on fields extracted
6466 from dates, where you often get leading zeros that should not be
6467 interpreted as octal.
6469 64. Added qualify_domain to the redirect router, to override the global
6472 65. If a pathologically long header line contained very many addresses (the
6473 report of this problem mentioned 10 000) and each of them was rewritten,
6474 Exim could use up a very large amount of memory. (It kept on making new
6475 copies of the header line as it rewrote, and never released the old ones.)
6476 At the expense of a bit more processing, the header rewriting function has
6477 been changed so that it no longer eats memory in this way.
6479 66. The generation of the Received: header has been moved from the time that a
6480 message starts to be received, to the time that it finishes. The timestamp
6481 in the Received: header should now be very close to that of the <= log
6482 line. There are two side-effects of this change:
6484 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6485 logged header lines no longer include the local Received: line, because
6486 it has not yet been created. The same applies to a copy of the message
6487 that is returned to a non-SMTP sender when a message is rejected.
6489 (b) When a filter file is tested using -bf, no additional Received: header
6490 is added to the test message. After some thought, I decided that this
6493 This change does not affect the value of $received_for. It is still set
6494 after address rewriting, but before local_scan() is called.
6496 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6498 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6499 gave an unhelpful panic error message, and a defer error. I have managed to
6500 change this behaviour so that it now rejects any supplied certificate,
6501 which seems right, as the list of acceptable certificates is empty.
6503 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6504 gave an unhelpful defer error. I have not managed to make this reject any
6505 supplied certificates, but the error message it gives is "no certificate
6506 supplied", which is not helpful.
6508 70. exigrep's output now also includes lines that are not associated with any
6509 message, but which match the given pattern. Implemented by a patch from
6510 Martin Sluka, which also tidied up the Perl a bit.
6512 71. Recipient callout verification, like sender verification, was using <> in
6513 the MAIL FROM command. This isn't really the right thing, since the actual
6514 sender may affect whether the remote host accepts the recipient or not. I
6515 have changed it to use the actual sender in the callout; this means that
6516 the cache record is now keyed on a recipient/sender pair, not just the
6517 recipient address. There doesn't seem to be a real danger of callout loops,
6518 since a callout by the remote host to check the sender would use <>.
6519 [SEE ABOVE: changed after hitting problems.]
6521 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6522 temporary errors. However, in the case of such a code being given after
6523 the end of a data transmission (i.e. after ".") Exim was failing to write
6524 a retry record for the message. (Yes, there was some broken host that was
6525 actually sending 8xx at this point.)
6527 73. An unknown lookup type in a host list could cause Exim to panic-die when
6528 the list was checked. (An example that provoked this was putting <; in the
6529 middle of a list instead of at the start.) If this happened during a DATA
6530 ACL check, a -D file could be left lying around. This kind of configuration
6531 error no longer causes Exim to die; instead it causes a defer error. The
6532 incident is still logged to the main and panic logs.
6534 74. Buglet left over from Exim 3 conversion. The message "too many messages
6535 in one connection" was written to the rejectlog but not the mainlog, except
6536 when address rewriting (yes!) was being logged.
6538 75. Added write_rejectlog option.
6540 76. When a system filter was run not as root (that is, when system_filter_user
6541 was set), the values of the $n variables were not being returned to the
6542 main process; thus, they were not subsequently available in the $sn
6545 77. Added +return_path_on_delivery log selector.
6547 78. A connection timeout was being treated differently from recipients deferred
6548 when testing hosts_max_try with a message that was older than the host's
6549 retry timeout. (The host should not be counted, thus allowing all hosts to
6550 be tried at least once before bouncing.) This may have been the cause of an
6551 occasionally reported bug whereby a message would remain on the queue
6552 longer than the retry timeout, but would be bounced if a delivery was
6553 forced. I say "may" because I never totally pinned down the problem;
6554 setting up timeout/retry tests is difficult. See also the next item.
6556 79. The ultimate address timeout was not being applied to errors that involved
6557 a combination of host plus message (for example, a timeout on a MAIL
6558 command). When an address resolved to a number of possible hosts, and they
6559 were not all tried for each delivery (e.g. because of hosts_max_try), a
6560 message could remain on the queue longer than the retry timeout.
6562 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6565 81. Fixed an obscure SMTP outgoing bug which required at least the following
6566 conditions: (a) there was another message waiting for the same server;
6567 (b) the server returned 5xx to all RCPT commands in the first message so
6568 that the message was not completed; (c) the server dropped the connection
6569 or gave a negative response to the RSET that Exim sends to abort the
6570 transaction. The observed case was a dropped connection after DATA that had
6571 been sent in pipelining mode. That is, the server had advertised PIPELINING
6572 but was not implementing it correctly. The effect of the bug was incorrect
6573 behaviour, such as trying another host, and this could lead to a crash.
6579 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6580 and daemon.c were passed as pointers to ints; they should have been
6581 pointers to socklen_t variables (which are typically unsigned ints).
6583 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6586 3. Fixed a really odd bug that affected only the testing scheme; patching a
6587 certain fixed string in the binary changed the value of another string that
6588 happened to be identical to the end of the original first string.
6590 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6591 name", it returns that address as the IP address. On some operating
6592 systems (e.g. Solaris), it also passes back the IP address string as the
6593 "host name". However, on others (e.g. Linux), it passes back an empty
6594 string. Exim wasn't checking for this, and was changing the host name to an
6595 empty string, assuming it had been canonicalized.
6597 5. Although rare, it is permitted to have more than one PTR record for a given
6598 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6599 all the names associated with an address, because they do in Solaris.
6600 However, it seems that they do not in Linux for data that comes from the
6601 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6602 I found this out when I moved to a new Linux workstation and tried to run
6603 the Exim test suite.
6605 To get round this problem I have changed the code so that it now does its
6606 own call to the DNS to look up PTR records when searching for a host name.
6607 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6608 addresses that are only in /etc/hosts are still found.
6610 This behaviour is, however, controlled by an option called host_lookup_
6611 order, which defaults to "bydns:byaddr". If people want to use the other
6612 order, or indeed, just use one or the other means of lookup, they can
6613 specify it in this variable.
6615 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6616 some operating systems, this comes back from gethostbyaddr() as an empty
6617 string, and this is what Exim used to test for. However, it seems that in
6618 other systems, "." is yielded. Exim now tests for this case too.
6620 7. The values of check_spool_space and check_log_space are now held internally
6621 as a number of kilobytes instead of an absolute number of bytes. If a
6622 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6623 kilobyte. This means that much larger values can be stored.
6625 8. Exim monitor: an attempt to get the action menu when not actually pointing
6626 at a message produces an empty menu entitled "No message selected". This
6627 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6628 no entries in it ("Shell widget menu has zero width and/or height"). So I
6629 have added a single, blank menu entry in this case.
6631 9. Added ${quote_local_part.
6633 10. MIME decoding is now applied to the contents of Subject: header lines when
6636 11. Now that a reference to $sender_host_address automatically causes a reverse
6637 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6638 host lookup before query-style lookups in lists that might use this
6639 variable. This has therefore been abolished, and the "net-" prefix is no
6640 longer necessary for query-style lookups.
6642 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6643 have been a typo for LFLAGS, so it has been changed.
6645 13. The install script calls Exim with "-C /dev/null" in order to find the
6646 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6647 to be output. However, since Exim outputs its version number before the
6648 error, it didn't break the script. It just looked ugly. I fixed this by
6649 always allowing "-C /dev/null" if the caller is root.
6651 14. Ignore overlarge ACL variable number when reading spool file - insurance
6652 against a later release with more variables having written the file.
6654 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6655 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6656 this, as well as the form without the "IPv6" on the front.
6658 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6659 OS/Makefile-Darwin file.
6661 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6662 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6664 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6665 variable. However, some OS use ints for inodes. I've added cast to long int
6666 to get rid of the compiler warning.
6668 19. I had forgotten to lock out "/../" in configuration file names when
6669 ALT_CONFIG_PREFIX was set.
6671 20. Routers used for verification do not need to specify transports. However,
6672 if such a router generated a host list, and callout was configured, Exim
6673 crashed, because it could not find a port number from the (non-existent)
6674 transport. It now assumes port 25 in this circumstance.
6676 21. Added the -t option to exigrep.
6678 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6679 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6680 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6681 them is set, LOOKUP_LSEARCH is forced.
6683 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6685 24. Added sender and host information to the "rejected by local_scan()" log
6686 line; previously there was no indication of these.
6688 25. Added .include_if_exists.
6690 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6691 Linux. It turns out that not all file systems support this. Apparently some
6692 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6693 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6694 not supported on the file descriptor, is now ignored when Exim is trying to
6695 sync a directory. This applies only to Linux.
6697 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6699 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6700 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6703 29. Added support for maildirsize files from supplied patch (modified a bit).
6705 30. The use of :fail: followed by an empty string could lead Exim to respond to
6706 sender verification failures with (e.g.):
6708 550 Verification failed for <xxx>
6709 550 Sender verify failed
6711 where the first response line was missing the '-' that indicates it is not
6712 the final line of the response.
6714 31. The loop for finding the name of the user that called Exim had a hardwired
6715 limit of 10; it now uses the value of finduser_retries, which is used for
6716 all other user lookups.
6718 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6721 33. Exim was neglecting to zero errno before one call of strtol() when
6722 expanding a string and expecting an integer value. On some systems this
6723 resulted in spurious "integer overflow" errors. Also, it was casting the
6724 result into an int without checking.
6726 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6727 did not work. The code looks as if it has *never* worked, though it appears
6728 to have been documented since at least release 1.62. I have made it work.
6730 35. The "timeout_DNS" error in retry rules, also documented since at least
6731 1.62, also never worked. As it isn't clear exactly what this means, and
6732 clearly it isn't a major issue, I have abolished the feature by treating it
6733 as "timeout", and writing a warning to the main and panic logs.
6735 36. The display of retry rules for -brt wasn't always showing the error code
6738 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6739 timeout_connect_A, timeout_connect_MX.
6741 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6742 to the empty sender.
6744 39. The daemon was not analysing the content of -oX till after it had closed
6745 stderr and disconnected from the controlling terminal. This meant that any
6746 syntax errors were only noted on the panic log, and the return code from
6747 the command was 0. By re-arranging the code a little, I've made the
6748 decoding happen first, so such errors now appear on stderr, and the return
6749 code is 1. However, the actual setting up of the sockets still happens in
6750 the disconnected process, so errors there are still only recorded on the
6753 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6754 connections (as happens on some IP stacks) was logged at start up time as
6755 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6756 it from "IPv6 and IPv4", which means that two separate sockets are being
6759 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6760 says whether AF_INET or AF_INET6 was passed as an argument.
6762 42. Exiwhat output was messed up when time zones were included in log
6765 43. Exiwhat now gives more information about the daemon's listening ports,
6766 and whether -tls-on-connect was used.
6768 44. The "port" option of the smtp transport is now expanded.
6770 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6771 silently ignored. Now an error message is written to the main and panic
6774 46. There's a new ACL modifier called "logwrite" which writes to a log file
6775 as soon as it is encountered.
6777 47. Added $local_user_uid and $local_user_gid at routing time.
6779 48. Exim crashed when trying to verify a sender address that was being
6782 49. Exim was recognizing only a space character after ".include". It now also
6783 recognizes a tab character.
6785 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6786 extracting the relevant information from the specification. The man page no
6787 longer contains scrambled data for the -d option, and I've added a section
6788 at the front about calling Exim under different names.
6790 51. Added "extra_headers" argument to the "mail" command in filter files.
6792 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6795 53. Installed eximstats 1.29.
6797 54. Added transport_filter_timeout as a generic transport option.
6799 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6800 Cc: header lines. This was required by RFC 822, but it not required by RFC
6803 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6804 incoming messages that did not have them. Now it does so only if the
6805 message originates locally, that is, if there is no associated remote host
6806 address. When Resent- header lines are present, this applies to the Resent-
6807 lines rather than the non-Resent- lines.
6809 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6810 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6812 58. Messages for configuration errors now include the name of the main
6813 configuration file - useful now that there may be more than one file in a
6814 list (.included file names were always shown).
6816 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6817 for those rare installations that do not start the daemon as root or run it
6818 setuid root. I've cut out the call to initgroups() if the daemon is not
6821 60. The Exim user and group can now be bound into the binary as text strings
6822 that are looked up at the start of Exim's processing.
6824 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6826 62. Added $mailstore_basename variable.
6828 63. Installed patch to sieve.c from Michael Haardt.
6830 64. When Exim failed to open the panic log after failing to open the main log,
6831 the original message it was trying to log was written to stderr and debug
6832 output, but if they were not available (the usual case in production), it
6833 was lost. Now it is written to syslog before the two lines that record the
6834 failures to open the logs.
6836 65. Users' Exim filters run in subprocesses under the user's uid. It is
6837 possible for a "deliver" command or an alias in a "personal" command to
6838 provoke an address rewrite. If logging of address rewriting is configured,
6839 this fails because the process is not running as root or exim. There may be
6840 a better way of dealing with this, but for the moment (because 4.30 needs
6841 to be released), I have disabled address rewrite logging when running a
6842 filter in a non-root, non-exim process.
6848 1. The buildconfig auxiliary program wasn't quoting the value set for
6849 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6850 not defined. This bug was masked in 4.22 by the effect that was fixed in
6853 2. Some messages that were rejected after a message id was allocated were
6854 shown as "incomplete" by exigrep. It no longer does this for messages that
6855 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6857 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6858 have allow_domain_literals set, the ID did not get logged in the <= line.
6859 Domain literals are now always recognized in Message-ID: header lines.
6861 4. The first argument for a ${extract expansion item is the key name or field
6862 number. Leading and trailing spaces in this item were not being ignored,
6863 causing some misleading effects.
6865 5. When deliver_drop_privilege was set, single queue runner processes started
6866 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6867 same command in the process it spins off) were not dropping privilege.
6869 6. When the daemon running as "exim" started a queue runner, it always
6870 re-executed Exim in the spun-off process. This is a waste of effort when
6871 deliver_drop_privilege is set. The new process now just calls the
6872 queue-runner function directly.
6878 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6881 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6882 ignored. Though the use of -oP was forcing the writing of a pid file, it
6883 was always written to the default place.
6885 3. If the message "no IP address found for host xxxx" is generated during
6886 incoming verification, it is now followed by identification of the incoming
6887 connection (so you can more easily find what provoked it).
6889 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6891 5. Added some features to "harden" Exim a bit more against certain attacks:
6893 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6894 be put in Local/Makefile. This is like the never_users runtime option,
6895 but it cannot be overridden. The default setting is "root".
6897 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6898 prefix string with which any file named in a -C command line option
6901 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6902 is retained for -C and -D only if the caller of Exim is root. Without
6903 it, the exim user may also use -C and -D and retain privilege.
6905 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6906 command line option is disabled.
6908 6. Macro names set by the -D option must start with an upper case letter, just
6909 like macro names defined in the configuration file.
6911 7. Added "dereference=" facility to LDAP.
6913 8. Two instances of the typo "uknown" in the source files are fixed.
6915 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6916 the Configure-Makefile script screwed up while processing it.
6918 10. Incorporated PCRE 4.4.
6920 11. The SMTP synchronization check was not operating right at the start of an
6921 SMTP session. For example, it could not catch a HELO sent before the client
6922 waited for the greeting. There is now a check for outstanding input at the
6923 point when the greeting is written. Because of the duplex, asynchronous
6924 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6925 way, but not yet received, when the check is performed.
6927 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6928 on TCP/IP sockets, because this apparently causes some broken clients to
6931 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6932 unchanged) from the Cygwin maintainer.
6934 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6935 was supported instead of testing for mbx. Effectively a typo.
6937 15. The spa authenticator server code was not checking that the input it
6938 received was valid base64.
6940 16. The debug output line for the "set" modifier in ACLs was not showing the
6941 name of the variable that was being set.
6943 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6945 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6946 lookup on its own. Something else was needed to trigger the lookup. For
6947 example, a match in host_lookup or the need for a host name in a host list.
6948 Now, if $sender_host_name is referenced and the host name has not yet been
6949 looked up, a lookup is performed. If the lookup fails, the variable remains
6950 empty, and $host_lookup_failed is set to "1".
6952 19. Added "eqi" as a case-independent comparison operator.
6954 20. The saslauthd authentication condition could segfault if neither service
6955 nor realm was specified.
6957 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6958 error message that was logged was misleading, and incoming SMTP
6959 connections were dropped. The message is now more accurate, and temporary
6960 errors are given to SMTP connections.
6962 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6963 (except RSET) with a 421 error, until QUIT is received. However, it was
6964 failing to send a response to QUIT.
6966 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6967 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6968 ACL is now run after the helo_try_verify_hosts code.
6970 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6971 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6972 case-independent, so other case variants are also recognized). Apparently
6973 some systems use these upper case variants.
6975 25. If more than two messages were waiting for the same host, and a transport
6976 filter was specified for the transport, Exim sent two messages over the
6977 same TCP/IP connection, and then failed with "socket operation on non-
6978 socket" when it tried to send the third.
6980 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6982 27. The extern definition of crypt16() in expand.c was not being excluded when
6983 the OS had its own crypt16() function.
6985 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6986 as a preferred synonym for return_size_limit, both as an option and as an
6989 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6991 30. Changed the default configuration ACL to relax the local part checking rule
6992 for addresses that are not in any local domains. For these addresses,
6993 slashes and pipe symbols are allowed within local parts, but the sequence
6994 /../ is explicitly forbidden.
6996 31. SPA server authentication was not clearing the challenge buffer before
6999 32. log_message in a "warn" ACL statement was writing to the reject log as
7000 well as to the main log, which contradicts the documentation and doesn't
7001 seem right (because no rejection is happening). So I have stopped it.
7003 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7004 However, I am unable to do any testing of this.
7006 34. Fixed an infelicity in the appendfile transport. When checking directories
7007 for a mailbox, to see if any needed to be created, it was accidentally
7008 using path names with one or more superfluous leading slashes; tracing
7009 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7011 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7012 contents are added to the log line that is written for every discarded
7013 recipient. (Previously a log_message setting was ignored.)
7015 36. The ${quote: operator now quotes the string if it is empty.
7017 37. The install script runs exim in order to find its version number. If for
7018 some reason other than non-existence or emptiness, which it checks, it
7019 could not run './exim', it was installing it with an empty version number,
7020 i.e. as "exim-". This error state is now caught, and the installation is
7023 38. An argument was missing from the function that creates an error message
7024 when Exim fails to connect to the socket for saslauthd authentication.
7025 This could cause Exim to crash, or give a corrupted message.
7027 39. Added isip, isip4, and isip6 to ${if conditions.
7029 40. The ACL variables $acl_xx are now saved with the message, and can be
7030 accessed later in routers, transports, and filters.
7032 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7033 strings in the file are not string-expanded.
7035 42. If a MAIL command specified a SIZE value that was too large to fit into an
7036 int variable, the check against message_size_limit failed. Such values are
7037 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7038 day this will have to be increased, but I don't think I want to be around
7039 when emails are that large.
7046 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7047 iconv() is not standard in FreeBSD.
7049 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7050 IPv6 enabled. The observed symptom was a segmentation fault on return from
7051 the function os_common_find_running_interfaces() in src/os.c.
7053 3. In the check_special_case() function in daemon.c I had used "errno" as an
7054 argument name, which causes warnings on some systems. This was basically a
7055 typo, since it was named "eno" in the comments!
7057 4. The code that waits for the clock to tick (at a resolution of some fraction
7058 of a second) so as to ensure message-id uniqueness was always waiting for
7059 at least one whole tick, when it could have waited for less. [This is
7060 almost certainly not relevant at current processor speeds, where it is
7061 unlikely to ever wait at all. But we try to future-proof.]
7063 5. The function that sleeps for a time interval that includes fractions of a
7064 second contained a race. It did not block SIGALRM between setting the
7065 timer, and suspending (a couple of lines later). If the interval was short
7066 and the sigsuspend() was delayed until after it had expired, the suspension
7067 never ended. On busy systems this could lead to processes getting stuck for
7070 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7071 process, before it forks any delivery processes. The open lookup caching
7072 mechanism meant that the open file or database connection was passed into
7073 the delivery process. The problem was that delivery processes always tidy
7074 up cached lookup data. This could cause a problem for the next delivery
7075 process started by the queue runner, because the external queue runner
7076 process does not know about the closure. So the next delivery process
7077 still has data in the lookup cache. In the case of a file lookup, there was
7078 no problem because closing a file descriptor in a subprocess doesn't affect
7079 the parent. However, if the lookup was caching a connection to a database,
7080 the connection was closed, and the second delivery process was likely to
7081 see errors such as "PGSQL: query failed: server closed the connection
7082 unexpectedly". The problem has been fixed by closing all cached lookups
7083 in a queue runner before running a delivery process.
7085 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7086 seem to have the "const" qualifier which it has on other OS. I've
7089 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7090 *of the same type* (client or server) with the same public name that an
7091 error should be diagnosed.
7093 9. When Exim looked up a host name for an IP address, but failed to find the
7094 original IP address when looking up the host name (a safety check), it
7095 output the message "<ip address> does not match any IP for NULL", which was
7096 confusing, to say the least. The bug was that the host name should have
7097 appeared instead of "NULL".
7099 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7100 user that is built into the binary, and the -C or -D options is used, root
7101 privilege is dropped before the configuration file is read. In addition,
7102 logging is switched to stderr instead of the normal log files. If the
7103 configuration then re-defines the Exim user, the unprivileged environment
7104 is probably not what is expected, so Exim logs a panic warning message (but
7107 However, if deliver_drop_privilege is set, the unprivileged state may well
7108 be exactly what is intended, so the warning has been cut out in that case,
7109 and Exim is allowed to try to write to its normal log files.
7115 1. smtp_return_error_details was not giving details for temporary sender
7116 or receiver verification errors.
7118 2. Diagnose a configuration error if two authenticators have the same public
7121 3. Exim used not to create the message log file for a message until the first
7122 delivery attempt. This could be confusing when incoming messages were held
7123 for policy or load reasons. The message log file is now created at the time
7124 the message is received, and an initial "Received" line is written to it.
7126 4. The automatically generated man page for command line options had a minor
7127 bug that caused no ill effects; however, a more serious problem was that
7128 the procedure for building the man page automatically didn't always
7129 operate. Consequently, release 4.20 contains an out-of-date version. This
7130 shouldn't happen again.
7132 5. When building Exim with embedded Perl support, the script that builds the
7133 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7134 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7136 6. The freeze_tell option was not being used for messages that were frozen on
7137 arrival, either by an ACL or by local_scan().
7139 7. Added the smtp_incomplete_transaction log selector.
7141 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7142 was accepting AUTH without a new EHLO.
7144 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7145 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7148 10. Logging of TCP/IP connections (when configured) now happens in the main
7149 daemon process instead of the child process, so that the TCP/IP connection
7150 count is more accurate (but it can never be perfect).
7152 11. The use of "drop" in a nested ACL was not being handled correctly in the
7153 outer ACL. Now, if condition failure induced by the nested "drop" causes
7154 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7155 or "require"), the connection is dropped.
7157 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7158 that yield "discard" can now be used with an "accept" or a "discard" verb,
7159 but an error is generated for any others (because I can't see a useful way
7160 to define what should happen).
7162 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7163 are now processed in the same way as lines in the Exim configuration file.
7164 In particular, continuation lines are supported.
7166 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7168 15. Added -ti meaning -t -i.
7170 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7171 domains, and warn by logging if others are found.
7173 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7174 returned from GIFCONF when Exim is trying to find the list of interfaces on
7175 a host. The code in os.c has been modified to copy each ifreq to an aligned
7176 structure in all cases.
7178 Also, in some cases, the returned ifreq's were being copied to a 'struct
7179 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7180 means the last couple of bytes of an IPv6 address could be chopped if the
7181 ifreq contained only a normal sockaddr (14 bytes storage).
7183 18. Named domain lists were not supported in the hosts_treat_as_local option.
7184 An entry such as +xxxx was not recognized, and was treated as a literal
7187 19. Ensure that header lines added by a DATA ACL are included in the reject log
7188 if the ACL subsequently rejects the message.
7190 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7191 MD5 (which is deprecated).
7193 21. When testing a filter file using -bf, Exim was writing a message when it
7194 took the sender from a "From " line in the message, but it was not doing so
7195 when it took $return_path from a Return-Path: header line. It now does.
7197 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7198 with a valid header line field name (a series of printing characters
7199 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7201 23. Changed "disc" in the source to "disk" to conform to the documentation and
7202 the book and for uniformity.
7204 24. Ignore Sendmail's -Ooption=value command line item.
7206 25. When execve() failed while trying to run a command in a pipe transport,
7207 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7208 could be confused with a return value of 69 from the command itself. This
7209 has been changed to 127, the value the shell returns if it is asked to run
7210 a non-existent command. The wording for the related log line suggests a
7211 non-existent command as the problem.
7213 26. If received_header_text expands to an empty string, do not add a Received:
7214 header line to the message. (Well, it adds a token one on the spool, but
7215 marks it "old" so that it doesn't get used or transmitted.)
7217 27. Installed eximstats 1.28 (addition of -nt option).
7219 28. There was no check for failure on the call to getsockname() in the daemon
7220 code. This can fail if there is a shortage of resources on the system, with
7221 ENOMEM, for example. A temporary error is now given on failure.
7223 29. Contrary to the C standard, it seems that in some environments, the
7224 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7225 program. Exim now does this explicitly; it affects the formatting of
7226 timestamps using strftime().
7228 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7229 complaints. I've now initialized all the variables, to avoid this.
7231 32. Header lines added by a system filter were not being "seen" during
7232 transport-time rewrites.
7234 33. The info_callback() function passed to OpenSSL is set up with type void
7235 (*)(SSL *, int, int), as described somewhere. However, when calling the
7236 function (actually a macro) that sets it up, the type void(*)() is
7237 expected. I've put in a cast to prevent warnings from picky compilers.
7239 34. If a DNS black list lookup found a CNAME record, but there were no A
7240 records associated with the domain it pointed at, Exim crashed.
7242 35. If a DNS black list lookup returned more than one A record, Exim ignored
7243 all but the first. It now scans all returned addresses if a particular IP
7244 value is being sought. In this situation, the contents of the
7245 $dnslist_value variable are a list of all the addresses, separated by a
7248 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7249 to accept a host name and all its aliases if the forward lookup for any of
7250 them yielded the IP address of the incoming connection. Now it accepts only
7251 those names whose forward lookup yields the correct IP address. Any other
7252 names are discarded. This closes a loophole whereby a rogue DNS
7253 administrator could create reverse DNS records to break through a
7254 wildcarded host restriction in an ACL.
7256 37. If a user filter or a system filter that ran in a subprocess used any of
7257 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7258 the wrong values were passed to the pipe command ($thisaddress had the
7259 value of $0, $0 had the value of $1, etc). This bug was introduced by
7260 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7262 38. Improved the line breaking for long SMTP error messages from ACLs.
7263 Previously, if there was no break point between 40 and 75 characters, Exim
7264 left the rest of the message alone. Two changes have been made: (a) I've
7265 reduced the minimum length to 35 characters; (b) if it can't find a break
7266 point between 35 and 75 characters, it looks ahead and uses the first one
7267 that it finds. This may give the occasional overlong line, but at least the
7268 remaining text gets split now.
7270 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7271 file descriptors might be low, and that setting 1000 would always raise it.
7272 It turns out that in some environments, the limit is already over 1000 and
7273 that lowering it causes trouble. So now Exim takes care not to decrease it.
7275 40. When delivering a message, the value of $return_path is set to $sender_
7276 address at the start of routing (routers may change the value). By an
7277 oversight, this default was not being set up when an address was tested by
7278 -bt or -bv, which affected the outcome if any router or filter referred to
7281 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7282 log only when "message" or "log_message" are set. However, if one of the
7283 conditions was an address verification, or a call to a nested ACL, the
7284 messages generated by the underlying test were being passed through. This
7285 no longer happens. The underlying message is available in $acl_verify_
7286 message for both "message" and "log_message" expansions, so it can be
7287 passed through if needed.
7289 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7290 new expansion $bh_ to give the encoded byte string without charset
7291 translation. Translation happens only if iconv() is available; HAVE_ICONV
7292 indicates this at build time. HEADERS_CHARSET gives the charset to
7293 translate to; headers_charset can change it in the configuration, and
7294 "headers charset" can change it in an individual filter file.
7296 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7297 that creates RFC 2047 encoded "words" labels them as that charset instead
7298 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7299 expansion operator; (ii) when Exim creates a From: line for a local
7300 message; (iii) when a header line is rewritten to include a "phrase" part.
7302 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7303 buggy, causing it to skip the first lines of messages whose message ID
7304 ended in 'D'. This would not have bitten before Exim release 4.14, because
7305 message IDs were unlikely to end in 'D' before then. The effect was to have
7306 incorrect size information for certain domains.
7308 45. #include "config.h" was missing at the start of the crypt16.c module. This
7309 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7312 46. If there was a timeout during a "random" callout check, Exim treated it as
7313 a failure of the random address, and carried on sending RSET and the real
7314 address. If the delay was just some slowness somewhere, the response to the
7315 original RCPT would be taken as a response to RSET and so on, causing
7316 mayhem of various kinds.
7318 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7319 when I implemented it. It didn't allow for the fact that some option values
7320 may legitimately be negative (e.g. size_addition), and it didn't even do
7321 the right test for positive values.
7323 48. Domain names in DNS records are case-independent. Exim always looks them up
7324 in lower case. Some resolvers return domain names in exactly the case they
7325 appear in the zone file, that is, they may contain uppercase letters. Not
7326 all resolvers do this - some return always lower case. Exim was treating a
7327 change of case by a resolver as a change of domain, similar to a widening
7328 of a domain abbreviation. This triggered its re-routing code and so it was
7329 trying to route what was effectively the same domain again. This normally
7330 caused routing to fail (because the router wouldn't handle the domain
7331 twice). Now Exim checks for this case specially, and just changes the
7332 casing of the domain that it ultimately uses when it transmits the message
7335 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7338 50. If a filter generated a file delivery with a non-absolute name (possible if
7339 no home directory exists for the router), the forbid_file option was not
7342 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7343 the existing equality matching.
7345 52. Exim was using ints instead of ino_t variables in some places where it was
7346 dealing with inode numbers.
7348 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7349 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7350 called TMPDIR, and if it finds it is different, it changes its value.
7352 54. The smtp_printf() function is now made available to local_scan() so
7353 additional output lines can be written before returning. There is also an
7354 smtp_fflush() function to enable the detection of a dropped connection.
7355 The variables smtp_input and smtp_batched_input are exported to
7358 55. Changed the default runtime configuration: the message "Unknown user"
7359 has been removed from the ACL, and instead placed on the localuser router,
7360 using the cannot_route_message feature. This means that any verification
7361 failures that generate their own messages won't get overridden. Similarly,
7362 the "Unrouteable address" message that was in the ACL for unverifiable
7363 relay addresses has also been removed.
7365 56. Added hosts_avoid_esmtp to the smtp transport.
7367 57. The exicyclog script was not checking for the esoteric option
7368 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7369 will work only if exicyclog is run under the appropriate euid.
7371 58. Following a discussion on the list, the rules by which Exim recognises line
7372 endings on incoming messages have been changed. The -dropcr and drop_cr
7373 options are now no-ops, retained only for backwards compatibility. The
7374 following line terminators are recognized: LF CRLF CR. However, special
7375 processing applies to CR:
7377 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7378 nor a local message in the state where . is a terminator.
7380 (ii) If a bare CR is encountered in a header line, an extra space is added
7381 after the line terminator so as not to end the header. The reasoning
7382 behind this is that bare CRs in header lines are most likely either
7383 to be mistakes, or people trying to play silly games.
7385 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7386 was being incorrectly given as 18 bytes larger than it should have been.
7387 This is a VOB (very old bug).
7389 60. This may never have affected anything current, but just in case it has:
7390 When the local host is found other than at the start of a list of hosts,
7391 the local host, those with the same MX, and any that follow, are discarded.
7392 When the list in question was part of a longer list of hosts, the following
7393 hosts (not currently being processed) were also being discarded. This no
7394 longer happens. I'm not sure if this situation could ever has previously
7397 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7400 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7401 header. This is something that is recommended in a new Internet Draft, and
7402 is something that is documented as being done by Sendmail. There are two
7403 possible values. For messages generated by the autoreply transport, Exim
7406 Auto-Submitted: auto-replied
7408 whereas for all other generated messages (e.g. bounces) it adds
7410 Auto-Submitted: auto-generated
7412 63. The "personal" condition in filters now includes a test for the
7413 Auto-Submitted: header. If it contains the string "auto-" the message it
7414 not considered personal.
7416 64. Added rcpt_include_affixes as a generic transport option.
7418 65. Added queue_only_override (default true).
7420 66. Added the syslog_duplication option.
7422 67. If what should have been the first header line of a message consisted of
7423 a space followed by a colon, Exim was mis-interpreting it as a header line.
7424 It isn't of course - it is syntactically invalid and should therefore be
7425 treated as the start of the message body. The misbehaviour could have
7426 caused a number of strange effects, including loss of data in subsequent
7427 header lines, and spool format errors.
7429 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7430 client host had authenticated. This control can now be exercised by an ACL
7431 for more flexibility.
7433 69. By default, callouts do not happen when testing with -bh. There is now a
7434 variant, -bhc, which does actually run the callout code, including
7435 consulting and updating the callout cache.
7437 70. Added support for saslauthd authentication, courtesy of Alexander
7440 71. If statvfs() failed on the spool or log directories while checking their
7441 size for availability, Exim confusingly gave the error "space shortage".
7442 Furthermore, in debugging mode it crashed with a floating point exception.
7443 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7444 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7445 serious problem, Exim now writes to the main and panic logs when this
7446 happens, with details of the failure. It then refuses to accept the
7447 incoming message, giving the message "spool directory problem" or "log
7448 directory problem" with a 421 code for SMTP messages.
7450 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7451 0, 1, and 2 exist, because some OS complain for execs without them (see
7452 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7453 descriptors. However, the code omitted to check that the open succeeded,
7454 causing mysterious errors if for some reason the permissions on /dev/null
7455 got screwed. Now Exim writes a message to the main and panic logs, and
7456 bombs out if it can't open /dev/null.
7458 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7459 interact so that it is all more flexible. It is supposed to remain
7460 backwards compatible. Also added extra_local_interfaces.
7462 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7463 to bomb out with an assertion failure - to the client this appears as a
7464 connection drop. This problem occurs in the part of the code that was taken
7465 from the Samba project. Fortunately, the assertion is in a very simple
7466 function, so I have fixed this by reproducing the function inline in the
7467 one place where it is called, and arranging for authentication to fail
7468 instead of killing the process with assert().
7470 75. The SPA client code was not working when the server requested OEM rather
7471 than Unicode encoding.
7473 76. Added code to make require_files with a specific uid setting more usable in
7474 the case where statting the file as root fails - usually a non-root-mounted
7475 NFS file system. When this happens and the failure is EACCES, Exim now
7476 forks a subprocess and does the per-uid checking as the relevant uid.
7478 77. Added process_log_path.
7480 78. If log_file_path was not explicitly set, a setting of check_log_space or
7481 check_log_inodes was ignored.
7483 79. If a space check for the spool or log partitions fails, the incident is now
7484 logged. Of course, in the latter case the data may get lost...
7486 80. Added the %p formatting code to string_format() so that it can be used to
7487 print addresses in debug_print(). Adjusted all the address printing in the
7488 debugging in store.c to use %p rather than %d.
7490 81. There was a concern that a line of code in smtp_in.c could overflow a
7491 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7492 initially expressed, the concern was not well-founded, because trailing
7493 spaces are removed early. However, if the trailing spaces were followed by
7494 a NULL, they did not get removed, so the overflow was possible. Two fixes
7497 (a) I re-wrote the offending code in a cleaner fashion.
7498 (b) If an incoming SMTP command contains a NULL character, it is rejected
7501 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7502 runs initgroups(), so that if the Exim user is in any additional groups,
7503 they will be used during message reception.
7509 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob