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.
125 JH/01 Remove code calling the customisable local_scan function, unless a new
126 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
128 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
129 non-signal-safe functions being used.
131 JH/03 Bug 2269: When presented with a received message having a stupidly large
132 number of DKIM-Signature headers, disable DKIM verification to avoid
133 a resource-consumption attack. The limit is set at twenty.
135 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
136 report of oldest_pass in ${authres } in consequence, and separate out
137 some descriptions of reasons for verification fail.
139 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
140 files in the spool were present and unlocked. A queue-runner could spot
141 them, resulting in a duplicate delivery. Fix that by doing the unlock
142 after the unlink. Investigation by Tim Stewart. Take the opportunity to
143 add more error-checking on spoolfile handling while that code is being
146 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
147 No known attacks, no CVE, this is defensive hardening.
149 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
150 a queue-runner could start a delivery while other operations were ongoing.
151 Cutthrough delivery was a common victim, resulting in duplicate delivery.
152 Found and investigated by Tim Stewart. Fix by using the open message data
153 file handle rather than opening another, and not locally closing it (which
154 releases a lock) for that case, while creating the temporary .eml format
155 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
157 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
158 $sender_verify_failure/$recipient_verify_failure to "random".
160 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
163 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
164 Previously this would segfault.
166 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
169 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
170 like zero, since the resolver should be doing this for us, But we need one
171 as a CNAME but no MX presence gets the CNAME returned; we need to check
172 that doesn't point to an MX to declare it "no MX returned" rather than
173 "error, loop". A new main option is added so the older capability of
174 following some limited number of chain links is maintained.
176 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
178 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
179 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
180 now-deprecated earlier definitions used only specified the range up to TLS
181 1.2 (in the older-version library docs).
183 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
185 JH/15 Rework TLS client-side context management. Stop using a global, and
186 explicitly pass a context around. This enables future use of TLS for
187 connections to service-daemons (eg. malware scanning) while a client smtp
188 connection is using TLS; with cutthrough connections this is quite likely.
190 JH/16 Fix ARC verification to do AS checks in reverse order.
192 JH/17 Support a "tls" option on the ${readsocket } expansion item.
194 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
195 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
196 Previously the "utf8" would be re-prepended for every additional message.
198 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
199 Previously thery were accepted, resulting in issues when attempting to
200 forward messages to a non-supporting MTA.
202 PP/02 Let -n work with printing macros too, not just options.
204 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
205 one parent address was copied, and bogus data was used at delivery-logging
206 time. Either a crash (after delivery) or bogus log data could result.
207 Discovery and analysis by Tim Stewart.
209 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
210 Previously if the string ended mid-character, we did not insert the
211 promised '?' replacement.
213 PP/04 Documentation: current string operators work on bytes, not codepoints.
215 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
216 should pack well giving a smaller memory footprint so better caching and
217 therefore performance. Group the declarations where this can't be done so
218 that the byte-sized flag variables are not interspersed among pointer
219 variables, giving a better chance of good packing by the compiler.
221 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
222 non-null, to avoid issues with sites running BATV. Previously reports were
223 sent with an empty envelope sender so looked like bounces.
225 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
226 The ignore_error flag wasn't being returned from the filter subprocess so
227 was not set for later routers. Investigation and fix by Matthias Kurz.
229 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
230 and a msg:complete for the whole, when a message is manually removed using
231 -Mrm. Developement by Matthias Kurz, hacked on by JH.
233 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
234 a "Gnu special" function, asprintf() in the DB utility binary builds; I
235 hope that is portable enough.
237 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
238 requiring a known-CA anchor certificate; make it now rely entirely on the
239 TLSA as an anchor. Checking the name on the leaf cert against the name
240 on the A-record for the host is still done for TA (but not for EE mode).
242 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
243 included in delivery lines for non-proxied connections, when compiled with
244 SUPPORT_SOCKS and running with proxy logging enabled.
246 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
247 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
248 move the existing event to fire before the normal logging of message
249 failure so that custom logging is bracketed by normal logging.
251 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
252 msg:fail:internal event. Developement by Matthias Kurz.
254 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
255 far too small for todays use of crypto signatures stored there. Go all
256 the way to the max DNS message size of 64kB, even though this might be
257 overmuch for IOT constrained device use.
259 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
260 copy a string over itself. The library routine is documented as not
261 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
263 JH/32 For main options check_spool_space and check_inode_space, where the
264 platform supports 64b integers, support more than the previous 2^31 kB
265 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
266 the previous G, M, k.
268 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
269 $authenticated_fail_id variable on authentication failure. Previously
272 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
273 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
274 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
275 GNUTLS_SEC_PARAM_MEDIUM.
277 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
278 side. Previously we would continue as if no SNI had been received.
280 JH/36 Harden the handling of string-lists. When a list consisted of a sole
281 "<" character, which should be a list-separator specification, we walked
282 off past the nul-terimation.
284 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
285 causes) even when the retry time is not yet met. Previously they were
286 not, meaning that when (say) an account was over-quota and temp-rejecting,
287 and multiple senders' messages were queued, only one sender would get
288 notified on each configured delay_warning cycle.
290 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
292 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
293 cipher-suites, an error can be left on the stack even for a succeeding
294 accept; this results in impossible error messages when a later operation
297 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
298 return error codes indicating retry. Under TLS1.3 this becomes required.
300 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
301 it only wrote the new authenticators, resulting in a lack of tracking of
302 peer changes of ESMTP extensions until the next cache flush.
304 JH/41 Fix the loop reading a message header line to check for integer overflow,
305 and more-often against header_maxsize. Previously a crafted message could
306 induce a crash of the recive process; now the message is cleanly rejected.
308 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
309 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
315 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
316 When redis_servers is set to a list of > 1 element, and the Redis servers
317 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
318 case of MOVED into a DEFER case instead, thus moving the query onto the
319 next server in the list. For a cluster of N elements, all N servers must
320 be defined in redis_servers.
322 GF/02 Catch and remove uninitialized value warning in exiqsumm
323 Check for existence of @ARGV before looking at $ARGV[0]
325 JH/01 Replace the store_release() internal interface with store_newblock(),
326 which internalises the check required to safely use the old one, plus
327 the allocate and data copy operations duplicated in both (!) of the
328 extant use locations.
330 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
331 modifier. This matches the restriction on the commandline.
333 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
334 Previously only the last row was returned.
336 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
337 we assumed that tags in the header were well-formed, and parsed the
338 element content after inspecting only the first char of the tag.
339 Assumptions at that stage could crash the receive process on malformed
342 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
343 While running the DKIM ACL we operate on the Permanent memory pool so that
344 variables created with "set" persist to the DATA ACL. Also (at any time)
345 DNS lookups that fail create cache records using the Permanent pool. But
346 expansions release any allocations made on the current pool - so a dnsdb
347 lookup expansion done in the DKIM ACL releases the memory used for the
348 DNS negative-cache, and bad things result. Solution is to switch to the
349 Main pool for expansions.
350 While we're in that code, add checks on the DNS cache during store_reset,
351 active in the testsuite.
352 Problem spotted, and debugging aided, by Wolfgang Breyha.
354 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
355 When none of the hosts presented to a transport match an already-open
356 connection, close it and proceed with the list. Previously we would
357 queue the message. Spotted by Lena with Yahoo, probably involving
360 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
361 Previously a spurious "250 OK id=" response was appended to the proper
364 JH/08 The "support for" informational output now, which built with Content
365 Scanning support, has a line for the malware scanner interfaces compiled
366 in. Interface can be individually included or not at build time.
368 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
369 by the template makefile "src/EDITME". The "STREAM" support for an older
370 ClamAV interface method is removed.
372 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
373 rows affected is given instead).
375 JH/11 The runtime Berkeley DB library version is now additionally output by
376 "exim -d -bV". Previously only the compile-time version was shown.
378 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
379 SMTP connection. Previously, when one had more receipients than the
380 first, an abortive onward connection was made. Move to full support for
381 multiple onward connections in sequence, handling cutthrough connection
382 for all multi-message initiating connections.
384 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
385 routers. Previously, a multi-recipient message would fail to match the
386 onward-connection opened for the first recipient, and cause its closure.
388 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
389 a timeout on read on a GnuTLS initiating connection, resulting in the
390 initiating connection being dropped. This mattered most when the callout
391 was marked defer_ok. Fix to keep the two timeout-detection methods
394 JH/15 Relax results from ACL control request to enable cutthrough, in
395 unsupported situations, from error to silently (except under debug)
396 ignoring. This covers use with PRDR, frozen messages, queue-only and
399 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
401 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
402 metadata, resulting in a crash in free().
404 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
405 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
406 Broken also in d185889f4, with init system revamp.
408 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
409 Previously we abruptly closed the connection after reading a malware-
410 found indication; now we go on to read the "scan ok" response line,
413 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
414 ACL. Previously, a crash would result.
416 JH/19 Speed up macro lookups during configuration file read, by skipping non-
417 macro text after a replacement (previously it was only once per line) and
418 by skipping builtin macros when searching for an uppercase lead character.
420 JH/20 DANE support moved from Experimental to mainline. The Makefile control
421 for the build is renamed.
423 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
424 was allocated for every new TLS startup, meaning one per message. Fix
425 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
427 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
428 reported the original. Fix to report (as far as possible) the ACL
429 result replacing the original.
431 JH/23 Fix memory leak during multi-message connections using STARTTLS under
432 OpenSSL. Certificate information is loaded for every new TLS startup,
433 and the resources needed to be freed.
435 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
437 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
440 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
441 DATA response info to the (existing) per-recipient response info for
442 the "C=" log element. It can have useful tracking info from the
443 destination system. Patch from Simon Arlott.
445 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
446 length value. Previously this would segfault.
448 HS/02 Support Avast multiline protoocol, this allows passing flags to
449 newer versions of the scanner.
451 JH/28 Ensure that variables possibly set during message acceptance are marked
452 dead before release of memory in the daemon loop. This stops complaints
453 about them when the debug_store option is enabled. Discovered specifically
454 for sender_rate_period, but applies to a whole set of variables.
455 Do the same for the queue-runner and queue-list loops, for variables set
456 from spool message files. Do the same for the SMTP per-message loop, for
457 certain variables indirectly set in ACL operations.
459 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
460 as a multi-recipient message from a mailinglist manager). The coding had
461 an arbitrary cutoff number of characters while checking for more input;
462 enforced by writing a NUL into the buffer. This corrupted long / fast
463 input. The problem was exposed more widely when more pipelineing of SMTP
464 responses was introduced, and one Exim system was feeding another.
465 The symptom is log complaints of SMTP syntax error (NUL chars) on the
466 receiving system, and refused recipients seen by the sending system
467 (propating to people being dropped from mailing lists).
468 Discovered and pinpointed by David Carter.
470 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
471 replaced by the ${authresults } expansion.
473 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
475 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
476 allows proper process termination in container environments.
478 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
479 Previously the "final dot" had a newline after it; ensure it is CR,LF.
481 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
482 and "err_perm", deprecated since 4.83 when the RFC-defined words
483 "temperror" and "permerror" were introduced.
485 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
486 transport-filters or DKIM-signing. The restriction was lost in the
487 consolidation of verify-callout and delivery SMTP handling.
488 Extend the restriction to also cover ARC-signing.
490 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
491 in defer=pass mode supply a 450 to the initiator. Previously the message
494 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
495 tls_require_ciphers is used as before.
497 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
498 "pass_unscanned". Only tmpfails from the scanner are written to
499 the paniclog, as they may require admin intervention (permission
500 denied, license issues). Other scanner errors (like decompression
501 bombs) do not cause a paniclog entry.
503 JH/36 Fix reinitialisation of DKIM logging variable between messages.
504 Previously it was possible to log spurious information in receive log
507 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
508 triggered odd behaviour from Outlook Express clients.
510 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
513 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
514 since the IETF WG has not yet settled on that versus the original
515 "bare" representation.
517 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
518 Previously the millisecond value corrupted the output.
519 Fix also for syslog_pid=no and log_selector +pid, for which the pid
520 corrupted the output.
526 JH/01 Rework error string handling in TLS interface so that the caller in
527 more cases is responsible for logging. This permits library-sourced
528 string to be attached to addresses during delivery, and collapses
529 pairs of long lines into single ones.
531 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
532 during configuration. Wildcards are allowed and expanded.
534 JH/02 Rework error string handling in DKIM to pass more info back to callers.
535 This permits better logging.
537 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
538 do not close it down and have the child transport start it up again on
539 the passed-on TCP connection. Instead, proxy the child (and any
540 subsequent ones) for TLS via a unix-domain socket channel. Logging is
541 affected: the continued delivery log lines do not have any DNSSEC, TLS
542 Certificate or OCSP information. TLS cipher information is still logged.
544 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
545 identical IP addresses on different listening ports. Will also affect
548 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
549 add noisy ifdef guards to special-case this sillyness.
550 Patch from Bernd Kuhls.
552 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
553 than 255 are no longer allowed.
555 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
556 Disable the session-cache too, which might reduce our load. Since we
557 currrectly use a new context for every connection, both as server and
558 client, there is no benefit for these.
559 GnuTLS appears to not support tickets server-side by default (we don't
560 call gnutls_session_ticket_enable_server()) but client side is enabled
561 by default on recent versions (3.1.3 +) unless the PFS priority string
564 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
565 <https://reproducible-builds.org/specs/source-date-epoch/>.
567 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
568 the check for any unsuccessful recipients did not notice the limit, and
569 erroneously found still-pending ones.
571 JH/08 Pipeline CHUNKING command and data together, on kernels that support
572 MSG_MORE. Only in-clear (not on TLS connections).
574 JH/09 Avoid using a temporary file during transport using dkim. Unless a
575 transport-filter is involved we can buffer the headers in memory for
576 creating the signature, and read the spool data file once for the
577 signature and again for transmission.
579 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
580 4.77 as the kernel support then wasn't solid, having issues in 64bit
581 mode. Now, it's been long enough. Add support for FreeBSD also.
583 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
584 case where the routing stage had gathered several addresses to send to
585 a host before calling the transport for the first, we previously failed
586 to close down TLS in the old transport process before passing the TCP
587 connection to the new process. The new one sent a STARTTLS command
588 which naturally failed, giving a failed delivery and bloating the retry
589 database. Investigation and fix prototype from Wolfgang Breyha.
591 JH/12 Fix check on SMTP command input synchronisation. Previously there were
592 false-negatives in the check that the sender had not preempted a response
593 or prompt from Exim (running as a server), due to that code's lack of
594 awareness of the SMTP input buffering.
596 PP/04 Add commandline_checks_require_admin option.
597 Exim drops privileges sanely, various checks such as -be aren't a
598 security problem, as long as you trust local users with access to their
599 own account. When invoked by services which pass untrusted data to
600 Exim, this might be an issue. Set this option in main configuration
601 AND make fixes to the calling application, such as using `--` to stop
604 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
605 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
606 responses to those, into a single TLS record each way (this usually means
607 a single packet). As a side issue, smtp_enforce_sync now works on TLS
610 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
611 affects you only if you're dancing at the edge of the param size limits.
612 If you are, and this message makes sense to you, then: raise the
613 configured limit or use OpenSSL 1.1. Nothing we can do for older
616 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
617 cmdline element to get the documented default one. Previously it was
620 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
621 are unsuable for non-message-specific reasons - eg. connection timeout,
622 banner-time rejection.
624 JH/16 Fix logging of delivery remote port, when specified by router, under
627 PP/06 Repair manualroute's ability to take options in any order, even if one
628 is the name of a transport.
631 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
633 JH/17 Change the list-building routines interface to use the expanding-string
634 triplet model, for better allocation and copying behaviour.
636 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
637 Previously it was constructed the first time a possibly-matching string
638 was met in the configuration file input during startup; now it is done
641 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
642 compatible one, to avoid the (poorly documented) possibility of a config
643 file in the working directory redirecting the DB files, possibly correpting
644 some existing file. CVE-2017-10140 assigned for BDB.
646 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
647 cache-hot. Previously, although the result was properly cached, the
648 initial verify call returned a defer.
650 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
651 the main verify for receipient in uncached-mode.
653 JH/22 Retire historical build files to an "unsupported" subdir. These are
654 defined as "ones for which we have no current evidence of testing".
656 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
657 if present. Previously it was ignored.
659 JH/24 Start using specified-initialisers in C structure init coding. This is
660 a C99 feature (it's 2017, so now considered safe).
662 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
663 if was a fixed-sized field and bitmask ops via macros; it is now more
666 PP/07 GitHub PR 56: Apply MariaDB build fix.
667 Patch provided by Jaroslav Škarvada.
669 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
670 during Coverity cleanups [4.87 JH/47]
671 Diagnosis and fix provided by Michael Fischer v. Mollard.
673 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
674 the right size to place the terminating semicolon on its own folded
675 line, the header hash was calculated to an incorrect value thanks to
676 the (relaxed) space the fold became.
678 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
679 and confused the parent.
681 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
682 which could crash as a result. This could lead to undeliverable messages.
684 JH/28 Logging: "next input sent too soon" now shows where input was truncated
687 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
688 matters on fast-turnover and PID-randomising systems, which were getting
689 out-of-order delivery.
691 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
692 a possibly-overlapping copy. The symptom was that "Remote host closed
693 connection in response to HELO" was logged instead of the actual 4xx
696 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
697 Previously only that bufferd was discarded, resulting in SYMTP command
700 JH/32 DKIM: when a message has multiple signatures matching an identity given
701 in dkim_verify_signers, run the dkim acl once for each. Previously only
702 one run was done. Bug 2189.
704 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
705 "panic the current process" to "deliberately defer". The panic log is
706 still written with the problem list name; the mail and reject logs now
707 get a temp-reject line for the message that was being handled, saying
708 something like "domains check lookup or other defer". The SMTP 451
709 message is still "Temporary local problem".
711 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
712 A crafted sequence of BDAT commands could result in in-use memory beeing
713 freed. CVE-2017-16943.
715 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
716 from SMTP input. Previously it was always done; now only done for DATA
717 and not BDAT commands. CVE-2017-16944.
719 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
720 to the message (such as an overlong header line). Previously this was
721 not done and we did not exit BDAT mode. Followon from the previous item
722 though a different problem.
728 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
729 than -2003 did; needs libidn2 in addition to libidn.
731 JH/02 The path option on a pipe transport is now expanded before use.
733 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
734 Patch provided by "Björn", documentation fix added too.
736 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
737 missing a wire-to-host endian conversion.
739 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
740 close after a BDAT command line could be taken as a following command,
741 giving a synch failure. Fix by only checking for synch immediately
742 before acknowledging the chunk.
744 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
745 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
746 macro. Patches provided by Josh Soref.
748 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
749 Previously we did not; the RFC seems ambiguous and VRFY is not listed
750 by IANA as a service extension. However, John Klensin suggests that we
753 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
754 the dkim code may be unix-mode line endings rather than smtp wire-format
755 CRLF, so prepend a CR to any bare LF.
757 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
758 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
760 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
761 the first were themselves being wrongly included in the feed into dkim
762 processing; with most chunk sizes in use this resulted in an incorrect
763 body hash calculated value.
765 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
766 DKIM signature block, for verification. Although advised against by
767 standards it is specifically not ruled illegal.
769 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
771 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
772 missing a body hash (the bh= tag).
774 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
775 It seems that HAProxy sends the Proxy Protocol information in clear and
776 only then does a TLS startup, so do the same.
778 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
779 TCP connections (such as for Spamd) unless the daemon successfully set
780 Fast Open mode on its listening sockets. This fixes breakage seen on
781 too-old kernels or those not configured for Fast Open, at the cost of
782 requiring both directions being enabled for TFO, and TFO never being used
783 by non-daemon-related Exim processes.
785 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
786 endings, at least on the first header line. Try to canonify any that get
787 past that check, despite the cost.
789 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
790 now limited to an arbitrary five deep, while parsing addresses with the
791 strip_excess_angle_brackets option enabled.
793 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
794 instead leave the unprompted TLS handshake in socket buffer for the
795 TLS library to consume.
797 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
799 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
801 JH/16 Drop variables when they go out of scope. Memory management drops a whole
802 region in one operation, for speed, and this leaves assigned pointers
803 dangling. Add checks run only under the testsuite which checks all
804 variables at a store-reset and panics on a dangling pointer; add code
805 explicitly nulling out all the variables discovered. Fixes one known
806 bug: a transport crash, where a dangling pointer for $sending_ip_address
807 originally assigned in a verify callout, is re-used.
809 PP/06 Drop '.' from @INC in various Perl scripts.
811 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
813 PP/08 Reduce a number of compilation warnings under clang; building with
814 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
815 should be warning-free.
817 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
819 HS/01 Fix portability problems introduced by PP/08 for platforms where
820 realloc(NULL) is not equivalent to malloc() [SunOS et al].
822 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
823 chunk. This allows us to accept broken chunked messages. We need a more
824 general solution here.
826 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
827 already-broken messages in the queue.
829 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
831 JH/19 Fix reference counting bug in routing-generated-address tracking.
837 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
838 supports it and a size is available (ie. the sending peer gave us one).
840 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
841 years of being deprecated). The replacements are the ACLs
842 acl_smtp_mime and acl_not_smtp_mime.
844 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
845 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
846 or even an in-clear connection were permitted. Now, if the host lookup
847 was dnssec and dane was requested then the host is only used if the
848 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
849 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
850 if one fails this test.
851 This means that a poorly-configured remote DNS will make it incommunicado;
852 but it protects against a DNS-interception attack on it.
854 JH/04 Bug 1810: make continued-use of an open smtp transport connection
855 non-noisy when a race steals the message being considered.
857 JH/05 If main configuration option tls_certificate is unset, generate a
858 self-signed certificate for inbound TLS connections.
860 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
861 in rewrites and routers.
863 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
864 and logged a warning sing 4.83; now they are a configuration file error.
866 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
867 (lacking @domain). Apply the same qualification processing as RCPT.
869 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
871 JH/10 Support ${sha256:} applied to a string (as well as the previous
874 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
875 a cutthrough deliver is pending, as we always want to make a connection.
876 This also avoids re-routing the message when later placing the cutthrough
877 connection after a verify cache hit.
878 Do not update it with the verify result either.
880 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
881 when routing results in more than one destination address.
883 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
884 signing (which inhibits the cutthrough capability). Previously only
885 the presence of an option was tested; now an expansion evaluating as
886 empty is permissible (obviously it should depend only on data available
887 when the cutthrough connection is made).
889 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
890 the relevant preceding SMTP command did not note the pipelining mode.
892 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
893 Previously they were not counted.
895 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
896 as one having no matching records. Previously we deferred the message
897 that needed the lookup.
899 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
900 distinguished as "(=".
902 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
903 for missing MX records. Previously it only worked for missing A records.
905 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
907 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
908 after the data-go-ahead and data-ack. Patch from Jason Betts.
910 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
911 even for a "none" policy. Patch from Tony Meyer.
913 JH/22 Fix continued use of a connection for further deliveries. If a port was
914 specified by a router, it must also match for the delivery to be
917 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
918 When one of the recipients of a message was unsuitable for the connection
919 (has no matching addresses), we lost track of needing to mark it
920 deferred. As a result mail would be lost.
922 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
924 JH/25 Decoding ACL controls is now done using a binary search; the source code
925 takes up less space and should be simpler to maintain. Merge the ACL
926 condition decode tables also, with similar effect.
928 JH/26 Fix problem with one_time used on a redirect router which returned the
929 parent address unchanged. A retry would see the parent address marked as
930 delivered, so not attempt the (identical) child. As a result mail would
933 JH/27 Fix a possible security hole, wherein a process operating with the Exim
934 UID can gain a root shell. Credit to http://www.halfdog.net/ for
935 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
938 JH/28 Enable {spool,log} filesystem space and inode checks as default.
939 Main config options check_{log,spool}_{inodes,space} are now
940 100 inodes, 10MB unless set otherwise in the configuration.
942 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
943 Previously it only applied to the main-section connection policy
946 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
948 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
949 by me. Added RFC7919 DH primes as an alternative.
951 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
952 are not in the system include path.
954 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
955 GnuTLS, when a session startup failed (eg because the client disconnected)
956 Exim did stdio operations after fclose. This was exposed by a recent
957 change which nulled out the file handle after the fclose.
959 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
960 signed directly by the cert-signing cert, rather than an intermediate
961 OCSP-signing cert. This is the model used by LetsEncrypt.
963 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
965 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
966 an incoming connection.
968 HS/02 Bug 1802: Do not half-close the connection after sending a request
971 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
972 fallback to "prime256v1".
974 JH/34 SECURITY: Use proper copy of DATA command in error message.
975 Could leak key material. Remotely exploitable. CVE-2016-9963.
981 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
982 and 3.4.4 - once the server is enabled to respond to an OCSP request
983 it does even when not requested, resulting in a stapling non-aware
984 client dropping the TLS connection.
986 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
987 support variable-length bit vectors. No functional change.
989 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
990 The I= interface field on outgoing lines is now after the H= remote
991 host field, same as incoming lines. There is a separate
992 outgoing_interface log selector which allows you to disable the
995 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
996 If not running log_selector +smtp_connection the mainlog would be held
997 open indefinitely after a "too many connections" event, including to a
998 deleted file after a log rotate. Leave the per net connection logging
999 leaving it open for efficiency as that will be quickly detected by the
1000 check on the next write.
1002 HS/01 Bug 1671: Fix post transport crash.
1003 Processing the wait-<transport> messages could crash the delivery
1004 process if the message IDs didn't exist for some reason. When
1005 using 'split_spool_directory=yes' the construction of the spool
1006 file name failed already, exposing the same netto behaviour.
1008 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1009 mime_regex ACL conditions.
1011 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1012 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1013 to HELO, local diagnostic string.
1015 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1016 log line to debug. Even when configured with a tls authenticator many
1017 client connections are expected to not authenticate in this way, so
1018 an authenticate fail is not an error.
1020 HS/02 Add the Exim version string to the process info. This way exiwhat
1021 gives some more detail about the running daemon.
1023 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1024 matter for fast-change records such as DNSBLs.
1026 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1027 retry record, even if constant. There may be multiple transports with
1028 different interface settings and the retry behaviour needs to be kept
1031 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1033 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1035 JH/10 Bug 840: fix log_defer_output option of pipe transport
1037 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1038 hosts_randomize. This matters a lot when combined with mua_wrapper.
1040 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1041 ${quote_pgsql:<string>} operator.
1043 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1045 JH/14 Change header file name for freeradius-client. Relevant if compiling
1046 with Radius support; from the Gentoo tree and checked under Fedora.
1048 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1050 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1051 option transport_filter, meaning no filtering.
1053 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1055 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1056 defaults to "*" (all hosts). The variable is now available when not built
1057 with TLS, default unset, mainly to enable keeping the testsuite sane.
1058 If a server certificate is not supplied (via tls_certificate) an error is
1059 logged, and clients will find TLS connections fail on startup. Presumably
1060 they will retry in-clear.
1061 Packagers of Exim are strongly encouraged to create a server certificate
1062 at installation time.
1064 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1065 with the $config_file variable.
1067 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1068 in transport context, after the attempt, and per-recipient. The latter type
1069 is per host attempted. The event data is the error message, and the errno
1070 information encodes the lookup type (A vs. MX) used for the (first) host,
1071 and the trailing two digits of the smtp 4xx response.
1073 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1074 to write to mainlog (or rejectlog, paniclog) in the window between file
1075 creation and permissions/ownership being changed. Particularly affects
1076 installations where exicyclog is run as root, rather than exim user;
1077 result is that the running daemon panics and dies.
1079 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1081 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1082 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1083 "pri" and "weight". Note that the previous implicit priority given by the
1084 list order is no longer honoured.
1086 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1087 for DKIM processing.
1089 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1090 by defining SUPPORT_SOCKS.
1092 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1093 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1094 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1095 variables are renamed to proxy_{local,external}_{address,port}.
1097 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1098 for a build by defining SUPPORT_I18N
1100 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1101 of the query string, and make ${quote_redis:} do that quoting.
1103 JH/29 Move Events support from Experimental to mainline, enabled by default
1104 and removable for a build by defining DISABLE_EVENT.
1106 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1108 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1109 cached by the daemon.
1111 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1112 by defining LOOKUP_REDIS. The libhiredis library is required.
1114 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1115 keys are given for lookup.
1117 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1118 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1119 only supported when built with TLS support. The PolarSSL SHA routines
1120 are still used when the TLS library is too old for convenient support.
1122 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1123 openssl_options), for security. OpenSSL forces this from version 1.1.0
1124 server-side so match that on older versions.
1126 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1127 allocation for $value could be released as the expansion processing
1128 concluded, but leaving the global pointer active for it.
1130 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1131 and to use the domains and local_parts ACL conditions.
1133 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1134 incorrectly not doubled on cutthrough transmission, hence seen as a
1135 body-termination at the receiving system - resulting in truncated mails.
1136 Commonly the sender saw a TCP-level error, and retransmitted the message
1137 via the normal store-and-forward channel. This could result in duplicates
1138 received - but deduplicating mailstores were liable to retain only the
1139 initial truncated version.
1141 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1143 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1145 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1146 we're in there, support oversigning also; bug 1309.
1148 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1150 HS/04 Add support for keep_environment and add_environment options.
1152 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1153 either intentional arithmetic overflow during PRNG, or testing config-
1156 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1157 delivery resulted in actual delivery. Cancel cutthrough before DATA
1160 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1161 rejects a recipient: pass the reject to the originator.
1163 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1164 Many were false-positives and ignorable, but it's worth fixing the
1167 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1168 for the new environment-manipulation done at startup. Move the routines
1169 from being local to tls.c to being global via the os.c file.
1171 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1172 an extract embedded as result-arg for a map, the first arg for extract
1173 is unavailable so we cannot tell if this is a numbered or keyed
1174 extraction. Accept either.
1180 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1183 JH/02 The smtp transport option "multi_domain" is now expanded.
1185 JH/03 The smtp transport now requests PRDR by default, if the server offers
1188 JH/04 Certificate name checking on server certificates, when exim is a client,
1189 is now done by default. The transport option tls_verify_cert_hostnames
1190 can be used to disable this per-host. The build option
1191 EXPERIMENTAL_CERTNAMES is withdrawn.
1193 JH/05 The value of the tls_verify_certificates smtp transport and main options
1194 default to the word "system" to access the system default CA bundle.
1195 For GnuTLS, only version 3.0.20 or later.
1197 JH/06 Verification of the server certificate for a TLS connection is now tried
1198 (but not required) by default. The verification status is now logged by
1199 default, for both outbound TLS and client-certificate supplying inbound
1202 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1205 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1206 Status Notification (bounce) messages are now MIME format per RFC 3464.
1207 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1208 under the control of the dsn_advertise_hosts option, and routers may
1209 have a dsn_lasthop option.
1211 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1212 default, modifiable by a malware= option. The list separator for
1213 the options can now be changed in the usual way. Bug 68.
1215 JH/10 The smtp_receive_timeout main option is now expanded before use.
1217 JH/11 The incoming_interface log option now also enables logging of the
1218 local interface on delivery outgoing connections.
1220 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1221 if the interface and destination host and port all match.
1223 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1226 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1227 Patch from Andrew Lewis.
1229 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1230 now supports optional time-restrictions, weighting, and priority
1231 modifiers per server. Patch originally by <rommer@active.by>.
1233 JH/16 The spamd_address main option now supports a mixed list of local
1234 and remote servers. Remote servers can be IPv6 addresses, and
1235 specify a port-range.
1237 JH/17 Bug 68: The spamd_address main option now supports an optional
1238 timeout value per server.
1240 JH/18 Bug 1581: Router and transport options headers_add/remove can
1241 now have the list separator specified.
1243 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1246 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1249 JH/21 Support for the A6 type of dns record is withdrawn.
1251 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1252 rather than the verbs used.
1254 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1255 from 255 to 1024 chars.
1257 JH/24 Verification callouts now attempt to use TLS by default.
1259 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1260 are generic router options now. The defaults didn't change.
1262 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1263 Original patch from Alexander Shikoff, worked over by JH.
1265 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1266 files if ZCAT_COMMAND is not executable.
1268 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1270 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1272 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1273 Normally benign, it bites when the pair was led to by a CNAME;
1274 modern usage is to not canonicalize the domain to a CNAME target
1275 (and we were inconsistent anyway for A-only vs AAAA+A).
1277 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1279 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1280 when evaluating $sender_host_dnssec.
1282 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1283 $sender_helo_dnssec variable.
1285 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1287 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1289 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1291 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1292 documented as working, but never had. Support all but $spam_report.
1294 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1295 added for tls authenticator.
1297 HS/03 Add perl_taintmode main config option
1303 TL/01 When running the test suite, the README says that variables such as
1304 no_msglog_check are global and can be placed anywhere in a specific
1305 test's script, however it was observed that placement needed to be near
1306 the beginning for it to behave that way. Changed the runtest perl
1307 script to read through the entire script once to detect and set these
1308 variables, reset to the beginning of the script, and then run through
1309 the script parsing/test process like normal.
1311 TL/02 The BSD's have an arc4random API. One of the functions to induce
1312 adding randomness was arc4random_stir(), but it has been removed in
1313 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1314 function when detected.
1316 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1317 cause callback expansion.
1319 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1320 syntax errors in an expansion can be treated as a string instead of
1321 logging or causing an error, due to the internal use of bool_lax
1322 instead of bool when processing it.
1324 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1325 server certificates when making smtp deliveries.
1327 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1329 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1331 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1333 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1334 Merged patch from Sebastian Wiedenroth.
1336 JH/05 Fix results-pipe from transport process. Several recipients, combined
1337 with certificate use, exposed issues where response data items split
1338 over buffer boundaries were not parsed properly. This eventually
1339 resulted in duplicates being sent. This issue only became common enough
1340 to notice due to the introduction of connection certificate information,
1341 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1343 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1344 size buffer was used, resulting in syntax errors when an expansion
1347 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1348 version 3.3.6 or later.
1350 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1351 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1352 both become "event_action", the variables become $event_name, $event_data
1353 and $event_defer_errno. There is a new variable $verify_mode, usable in
1354 routers, transports and related events. The tls:cert event is now also
1355 raised for inbound connections, if the main configuration event_action
1358 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1359 early OCSP support, but no stapling (appears to be less than 1.0.0).
1361 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1362 server certificate names available under the smtp transport option
1363 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1366 JH/10 Time-related extraction expansions from certificates now use the main
1367 option "timezone" setting for output formatting, and are consistent
1368 between OpenSSL and GnuTLS compilations. Bug 1541.
1370 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1371 encoded parameter in the incoming message. Bug 1558.
1373 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1374 include certificate info, eximon was claiming there were spoolfile
1377 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1379 JH/14 Log delivery-related information more consistently, using the sequence
1380 "H=<name> [<ip>]" wherever possible.
1382 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1383 are problematic for Debian distribution, omit them from the release
1386 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1388 JH/16 Fix string representation of time values on 64bit time_t architectures.
1391 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1392 output list separator was used.
1397 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1398 checkers that were complaining about end of non-void function with no
1401 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1402 This was a regression introduced in 4.83 by another bugfix.
1404 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1406 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1407 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1413 TF/01 Correctly close the server side of TLS when forking for delivery.
1415 When a message was received over SMTP with TLS, Exim failed to clear up
1416 the incoming connection properly after forking off the child process to
1417 deliver the message. In some situations the subsequent outgoing
1418 delivery connection happened to have the same fd number as the incoming
1419 connection previously had. Exim would try to use TLS and fail, logging
1420 a "Bad file descriptor" error.
1422 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1423 utilities have not been installed.
1425 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1426 temporary space as the ACL may create new global variables.
1428 TL/01 LDAP support uses per connection or global context settings, depending
1429 upon the detected version of the libraries at build time.
1431 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1432 to extract and use the src ip:port in logging and expansions as if it
1433 were a direct connection from the outside internet. PPv2 support was
1434 updated based on HAProxy spec change in May 2014.
1436 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1438 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1439 Properly escape header and check for NULL return.
1441 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1444 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1446 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1447 characters in header names, implemented as a verify condition.
1448 Contributed by Michael Fischer v. Mollard.
1450 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1451 results permerror and temperror. Previous values are deprecated but
1452 still accepted. In a future release, err_perm and err_temp will be
1453 completely removed, which will be a backward incompatibility if the
1454 ACL tests for either of these two old results. Patch contributed by
1455 user bes-internal on the mailing list.
1457 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1459 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1460 selectors, in both main and reject logs.
1462 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1465 JH/07 Add malware type "sock" for talking to simple daemon.
1467 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1469 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1470 routers/transports under cutthrough routing.
1472 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1473 numbers. Touch up "bool" conditional to keep the same definition.
1475 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1477 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1479 JH/12 Expand items in router/transport headers_add or headers_remove lists
1480 individually rather than the list as a whole. Bug 1452.
1482 Required for reasonable handling of multiple headers_ options when
1483 they may be empty; requires that headers_remove items with embedded
1484 colons must have them doubled (or the list-separator changed).
1486 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1487 view the policy declared in the DMARC record. Currently, $dmarc_status
1488 is a combined value of both the record presence and the result of the
1491 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1493 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1494 dnslookup router and the smtp transport (applying to the forward
1497 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1498 of ldap servers used for a specific lookup. Patch provided by Heiko
1501 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1502 New variable $lookup_dnssec_authenticated for observability.
1504 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1505 Patch submitted by Lars Timman.
1507 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1509 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1510 Requires trusted mode and valid format message id, aborts otherwise.
1511 Patch contributed by Heiko Schlichting.
1513 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1514 certextract with support for various fields. Bug 1358.
1516 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1517 is requested by default, modifiable by smtp transport option
1520 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1521 operate on certificate variables to give certificate fingerprints
1522 Also new ${sha256:cert_variable}.
1524 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1526 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1527 Christian Aistleitner.
1529 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1531 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1532 file. Patch from Wolfgang Breyha.
1534 JH/25 Expand the coverage of the delivery $host and $host_address to
1535 client authenticators run in verify callout. Bug 1476.
1537 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1538 align with daemon_smtp_ports. Bug 72.
1540 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1541 support and error reporting did not work properly.
1543 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1544 and is readable. Patch from Andrew Colin Kissa.
1546 TL/14 Enhance documentation of ${run expansion and how it parses the
1547 commandline after expansion, particularly in the case when an
1548 unquoted variable expansion results in an empty value.
1550 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1552 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1553 by renaming away from it. Addresses GH issue 15, reported by
1556 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1558 TL/15 SECURITY: prevent double expansion in math comparison functions
1559 (can expand unsanitized data). Not remotely exploitable.
1566 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1568 PP/02 Make -n do something, by making it not do something.
1569 When combined with -bP, the name of an option is not output.
1571 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1574 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1575 $sender_host_name and config options to manage this, and basic check
1578 PP/05 DSCP support for outbound connections and control modifier for inbound.
1580 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1581 (Only plugin which currently uses this is kerberos4, which nobody should
1582 be using, but we should make it available and other future plugins might
1583 conceivably use it, even though it would break NAT; stuff *should* be
1584 using channel bindings instead).
1586 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1587 name; added for Sendmail compatibility; requires admin caller.
1588 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1589 just ignore it); requires trusted caller.
1590 Also parse but ignore: -Ac -Am -X<logfile>
1593 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1595 TL/02 Add +smtp_confirmation as a default logging option.
1597 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1598 Patch by Magnus Holmgren from 2007-02-20.
1600 TL/04 Bugzilla 1281 - Spec typo.
1601 Bugzilla 1283 - Spec typo.
1602 Bugzilla 1290 - Spec grammar fixes.
1604 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1606 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1608 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1609 mailing list by by Dmitry Isaikin.
1611 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1613 JH/02 Support "G" suffix to numbers in ${if comparisons.
1615 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1617 NM/01 Bugzilla 1197 - Spec typo
1618 Bugzilla 1196 - Spec examples corrections
1620 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1622 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1623 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1626 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1627 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1629 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1630 "acl {{name}{arg}...}", and optional args on acl condition
1633 JH/05 Permit multiple router/transport headers_add/remove lines.
1635 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1637 JH/07 Avoid using a waiting database for a single-message-only transport.
1638 Performance patch from Paul Fisher. Bugzilla 1262.
1640 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1643 JH/09 Add $headers_added variable, with content from use of ACL modifier
1644 add_header (but not yet added to the message). Bugzilla 199.
1646 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1647 Pulled from Bugzilla 817 by Wolfgang Breyha.
1649 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1651 (nb: this is the same fix as in Exim 4.80.1)
1653 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1656 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1659 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1661 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1662 advertise SMTP AUTH mechanism to us, instead of a generic
1663 protocol violation error. Also, make Exim more robust to bad
1664 data from the Dovecot auth socket.
1666 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1668 When a queue runner is handling a message, Exim first routes the
1669 recipient addresses, during which it prunes them based on the retry
1670 hints database. After that it attempts to deliver the message to
1671 any remaining recipients. It then updates the hints database using
1674 So if a recipient address works intermittently, it can get repeatedly
1675 deferred at routing time. The retry hints record remains fresh so the
1676 address never reaches the final cutoff time.
1678 This is a fairly common occurrence when a user is bumping up against
1679 their storage quota. Exim had some logic in its local delivery code
1680 to deal with this. However it did not apply to per-recipient defers
1681 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1683 This change adds a proper retry rule check during routing so that the
1684 final cutoff time is checked against the message's age. We only do
1685 this check if there is an address retry record and there is not a
1686 domain retry record; this implies that previous attempts to handle
1687 the address had the retry_use_local_parts option turned on. We use
1688 this as an approximation for the destination being like a local
1689 delivery, as in LMTP.
1691 I suspect this new check makes the old local delivery cutoff check
1692 redundant, but I have not verified this so I left the code in place.
1694 TF/02 Correct gecos expansion when From: is a prefix of the username.
1696 Test 0254 submits a message to Exim with the header
1700 When I ran the test suite under the user fanf2, Exim expanded
1701 the header to contain my full name, whereas it should have added
1702 a Resent-Sender: header. It erroneously treats any prefix of the
1703 username as equal to the username.
1705 This change corrects that bug.
1707 GF/01 DCC debug and logging tidyup
1708 Error conditions log to paniclog rather than rejectlog.
1709 Debug lines prefixed by "DCC: " to remove any ambiguity.
1711 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1713 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1714 Bug spotted by Jeremy Harris; was flawed since initial commit.
1715 Would have resulted in OCSP responses post-SNI triggering an Exim
1716 NULL dereference and crash.
1718 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1720 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1721 Bug detection, analysis and fix by Samuel Thibault.
1722 Bugzilla 1331, Debian bug #698092.
1724 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1726 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1727 Server implementation by Todd Lyons, client by JH.
1728 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1729 config variable "prdr_enable" controls whether the server
1730 advertises the facility. If the client requests PRDR a new
1731 acl_data_smtp_prdr ACL is called once for each recipient, after
1732 the body content is received and before the acl_smtp_data ACL.
1733 The client is controlled by both of: a hosts_try_prdr option
1734 on the smtp transport, and the server advertisement.
1735 Default client logging of deliveries and rejections involving
1736 PRDR are flagged with the string "PRDR".
1738 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1739 fclose(). Diagnosis by Todd Lyons.
1741 PP/17 Update configure.default to handle IPv6 localhost better.
1742 Patch by Alain Williams (plus minor tweaks).
1745 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1746 This is now consistent with GnuTLS, and is now documented: the
1747 previous undocumented portable approach to treating the option as
1748 unset was to force an expansion failure. That still works, and
1749 an empty string is now equivalent.
1751 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1752 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1753 not performing validation itself.
1755 PP/20 Added force_command boolean option to pipe transport.
1756 Patch from Nick Koston, of cPanel Inc.
1758 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1761 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1763 PP/21 Fix eximon continuous updating with timestamped log-files.
1764 Broken in a format-string cleanup in 4.80, missed when I repaired the
1765 other false fix of the same issue.
1766 Report and fix from Heiko Schlichting.
1769 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1770 Report from Prashanth Katuri.
1772 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1773 It's SecureTransport, so affects any MacOS clients which use the
1774 system-integrated TLS libraries, including email clients.
1776 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1777 using a MIME ACL for non-SMTP local injection.
1778 Report and assistance in diagnosis by Warren Baker.
1780 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1782 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1784 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1785 last id that failed so it may be referenced in subsequent ACL's.
1787 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1790 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1791 ldap library initialization, allowing self-signed CA's to be
1792 used. Also properly sets require_cert option later in code by
1793 using NULL (global ldap config) instead of ldap handle (per
1794 session). Bug diagnosis and testing by alxgomz.
1796 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1797 the src/util/ subdirectory.
1799 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1800 renamed to Transport Post Delivery Action by Jeremy Harris, as
1803 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1804 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1805 redis_servers = needs to be configured which will be used by the redis
1806 lookup. Patch from Warren Baker, of The Packet Hub.
1808 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1810 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1811 hostname or reverse DNS when processing a host list. Used suggestions
1812 from multiple comments on this bug.
1814 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1816 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1817 Missed a few lines, added it to make the runtest require no keyboard
1820 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1821 contains upper case chars. Make router use caseful_local_part.
1823 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1824 support when GnuTLS has been built with p11-kit.
1830 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1832 This, or similar/improved, will also be change PP/11 of 4.82.
1838 PP/01 Handle short writes when writing local log-files.
1839 In practice, only affects FreeBSD (8 onwards).
1840 Bugzilla 1053, with thanks to Dmitry Isaikin.
1842 NM/01 Bugzilla 949 - Documentation tweak
1844 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1847 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1849 PP/02 Implemented gsasl authenticator.
1851 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1853 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1854 `pkg-config foo` for cflags/libs.
1856 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1857 with rest of GSASL and with heimdal_gssapi.
1859 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1860 `pkg-config foo` for cflags/libs for the TLS implementation.
1862 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1863 properties get this fed in as external SSF. A number of robustness
1864 and debugging improvements to the cyrus_sasl authenticator.
1866 PP/08 cyrus_sasl server now expands the server_realm option.
1868 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1869 Patch by Jeremy Harris.
1871 PP/10 Added dbmjz lookup type.
1873 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1875 PP/12 MAIL args handles TAB as well as SP, for better interop with
1876 non-compliant senders.
1877 Analysis and variant patch by Todd Lyons.
1879 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1880 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1881 Patch from Dirk Mueller <dmueller@suse.com>
1883 PP/13 tls_peerdn now print-escaped for spool files.
1884 Observed some $tls_peerdn in wild which contained \n, which resulted
1885 in spool file corruption.
1887 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1888 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1889 or write after TLS renegotiation, which otherwise led to messages
1892 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1893 as a tracking header (ie: a signed header comes before the signature).
1894 Patch from Wolfgang Breyha.
1896 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1897 comma-sep list; embedded commas doubled.
1899 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1901 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1903 Report and patch from Dmitry Banschikov.
1905 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1906 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1907 failures. We appear to now support TLS1.1+ with Exim.
1909 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1910 lets Exim select keys and certificates based upon TLS SNI from client.
1911 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1912 before an outbound SMTP session. New log_selector, +tls_sni.
1914 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1915 NULL dereference. Report and patch from Alun Jones.
1917 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1918 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1919 Not seeing resolver debug output on NetBSD, but suspect this is a
1920 resolver implementation change.
1922 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1923 Left warnings. Added "eximon gdb" invocation mode.
1925 PP/21 Defaulting "accept_8bitmime" to true, not false.
1927 PP/22 Added -bw for inetd wait mode support.
1929 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1930 locate the relevant includes and libraries. Made this the default.
1932 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1933 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1935 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1936 This may cause build issues on older platforms.
1938 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1939 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1940 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1941 Added SNI support via GnuTLS too.
1942 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1944 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1946 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1947 Applied second patch from Janne, implementing suggestion to default
1948 multiple-strings-in-record handling to match SPF spec.
1950 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1952 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1953 read-only, out of scope).
1954 Patch from Wolfgang Breyha, report from Stuart Northfield.
1956 PP/29 Fix three issues highlighted by clang analyser static analysis.
1957 Only crash-plausible issue would require the Cambridge-specific
1958 iplookup router and a misconfiguration.
1959 Report from Marcin Mirosław.
1961 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1963 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1964 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1965 As part of this, removing so much warning spew let me fix some minor
1966 real issues in debug logging.
1968 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1969 assignment on my part. Fixed.
1971 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1972 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1973 Janne Snabb (who went above and beyond: thank you).
1975 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1976 string otherwise requires a connection and a bunch more work and it's
1977 relatively easy to get wrong. Should also expose TLS library linkage
1980 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1981 64-bit ${eval} (JH/03).
1983 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1984 GNU libc to support some of the 64-bit stuff, should not lead to
1985 conflicts. Defined before os.h is pulled in, so if a given platform
1986 needs to override this, it can.
1988 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1989 protection layer was required, which is not implemented.
1990 Bugzilla 1254, patch from Wolfgang Breyha.
1992 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1993 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1994 tls_dhparam take prime identifiers. Also unbreak combination of
1995 OpenSSL+DH_params+TLSSNI.
1997 PP/39 Disable SSLv2 by default in OpenSSL support.
2003 PP/01 Solaris build fix for Oracle's LDAP libraries.
2004 Bugzilla 1109, patch from Stephen Usher.
2006 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2008 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2011 TF/02 Fix a couple more cases where we did not log the error message
2012 when unlink() failed. See also change 4.74-TF/03.
2014 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2015 lock up or crash if it happened to be inside a call to libc when it
2016 got a SIGUSR1 from exiwhat.
2018 The SIGUSR1 handler appends the current process status to the process
2019 log which is later printed by exiwhat. It used to use the general
2020 purpose logging code to do this, but several functions it calls are
2021 not safe for signals.
2023 The new output code in the SIGUSR1 handler is specific to the process
2024 log, and simple enough that it's easy to inspect for signal safety.
2025 Removing some special cases also simplifies the general logging code.
2026 Removing the spurious timestamps from the process log simplifies
2029 TF/04 Improved ratelimit ACL condition.
2031 The /noupdate option has been deprecated in favour of /readonly which
2032 has clearer semantics. The /leaky, /strict, and /readonly update modes
2033 are mutually exclusive. The update mode is no longer included in the
2034 database key; it just determines when the database is updated. (This
2035 means that when you upgrade Exim will forget old rate measurements.)
2037 Exim now checks that the per_* options are used with an update mode that
2038 makes sense for the current ACL. For example, when Exim is processing a
2039 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2040 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2041 must specify per_mail/readonly. If you omit the update mode it defaults to
2042 /leaky where that makes sense (as before) or /readonly where required.
2044 The /noupdate option is now undocumented but still supported for
2045 backwards compatibility. It is equivalent to /readonly except that in
2046 ACLs where /readonly is required you may specify /leaky/noupdate or
2047 /strict/noupdate which are treated the same as /readonly.
2049 A useful new feature is the /count= option. This is a generalization
2050 of the per_byte option, so that you can measure the throughput of other
2051 aggregate values. For example, the per_byte option is now equivalent
2052 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2054 The per_rcpt option has been generalized using the /count= mechanism
2055 (though it's more complicated than the per_byte equivalence). When it is
2056 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2057 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2058 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2059 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2060 Note that using per_rcpt with a non-readonly update mode in more than
2061 one ACL will cause the recipients to be double-counted. (The per_mail
2062 and per_byte options don't have this problem.)
2064 The handling of very low rates has changed slightly. If the computed rate
2065 is less than the event's count (usually one) then this event is the first
2066 after a long gap. In this case the rate is set to the same as this event's
2067 count, so that the first message of a spam run is counted properly.
2069 The major new feature is a mechanism for counting the rate of unique
2070 events. The new per_addr option counts the number of different
2071 recipients that someone has sent messages to in the last time period. It
2072 behaves like per_rcpt if all the recipient addresses are different, but
2073 duplicate recipient addresses do not increase the measured rate. Like
2074 the /count= option this is a general mechanism, so the per_addr option
2075 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2076 example, measure the rate that a client uses different sender addresses
2077 with the options per_mail/unique=$sender_address. There are further
2078 details in the main documentation.
2080 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2082 TF/06 Removed a few PCRE remnants.
2084 TF/07 Automatically extract Exim's version number from tags in the git
2085 repository when doing development or release builds.
2087 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2088 Bugzilla 879. Patch from Paul Fisher.
2090 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2091 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2094 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2096 PP/05 Variable $av_failed, true if the AV scanner deferred.
2097 Bugzilla 1078. Patch from John Horne.
2099 PP/06 Stop make process more reliably on build failure.
2100 Bugzilla 1087. Patch from Heiko Schlittermann.
2102 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2103 Bugzilla 1089. Patch from Heiko Schlittermann.
2105 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2106 Bugzilla 1131. Patch from Holger Weiß.
2108 PP/09 Handle IPv6 addresses with SPF.
2109 Bugzilla 860. Patch from Wolfgang Breyha.
2111 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2113 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2116 PP/11 match_* no longer expand right-hand-side by default.
2117 New compile-time build option, EXPAND_LISTMATCH_RHS.
2118 New expansion conditions, "inlist", "inlisti".
2120 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2122 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2124 PP/14 fix log_write() format string regression from TF/03.
2125 Bugzilla 1152. Patch from Dmitry Isaikin.
2131 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2133 PP/02 Harmonised TLS library version reporting; only show if debugging.
2134 Layout now matches that introduced for other libraries in 4.74 PP/03.
2136 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2138 PP/04 New "dns_use_edns0" global option.
2140 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2143 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2144 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2146 TK/01 Updated PolarSSL code to 0.14.2.
2147 Bugzilla 1097. Patch from Andreas Metzler.
2149 PP/07 Catch divide-by-zero in ${eval:...}.
2150 Fixes bugzilla 1102.
2152 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2155 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2156 format-string attack -- SECURITY: remote arbitrary code execution.
2158 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2159 time unintentionally subject to list matching rules, letting the header
2160 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2161 arbitrary string expansion). This allowed for information disclosure.
2163 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2164 INT_MIN/-1 -- value coerced to INT_MAX.
2170 NM/01 Workaround for PCRE version dependency in version reporting
2173 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2174 This fixes portability to compilers other than gcc, notably
2175 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2177 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2178 makefiles for portability to HP-UX and POSIX correctness.
2180 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2181 Also via indented variable definition in the Makefile.
2182 (Debugging by Oliver Heesakkers).
2184 PP/02 Restore caching of spamd results with expanded spamd_address.
2185 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2187 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2188 Improves build reliability. Fix from: Frank Elsner
2190 NM/02 Fix wide character breakage in the rfc2047 coding
2191 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2193 NM/03 Allow underscore in dnslist lookups
2194 Fixes bug 1026. Patch from Graeme Fowler
2196 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2197 Code patches from Adam Ciarcinski of NetBSD.
2199 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2202 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2203 is logged, to avoid truncation. Patch from John Horne.
2205 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2206 Patch from Jakob Hirsch.
2208 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2209 SQL string expansion failure details.
2210 Patch from Andrey Oktyabrski.
2212 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2213 Patch from Simon Arlott.
2215 PP/09 New lookups functionality failed to compile on old gcc which rejects
2216 extern declarations in function scope.
2217 Patch from Oliver Fleischmann
2219 PP/10 Use sig_atomic_t for flags set from signal handlers.
2220 Check getgroups() return and improve debugging.
2221 Fixed developed for diagnosis in bug 927 (which turned out to be
2224 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2225 Patch from Mark Zealey.
2227 PP/12 Bugzilla 1056: Improved spamd server selection.
2228 Patch from Mark Zealey.
2230 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2231 Based on patch from Heiko Schlittermann.
2233 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2234 Patch from Uwe Doering, sign-off by Michael Haardt.
2236 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2237 variable declaration deep within a block. Bug and patch from
2240 PP/15 lookups-Makefile IRIX compatibility coercion.
2242 PP/16 Make DISABLE_DKIM build knob functional.
2244 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2245 Patch by Simon Arlott
2247 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2248 variable argument macros. Our copy now differs from upstream.
2254 TF/01 Failure to get a lock on a hints database can have serious
2255 consequences so log it to the panic log.
2257 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2258 controlled using the smtp_confirmation log selector.
2260 TF/03 Include the error message when we fail to unlink a spool file.
2262 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2263 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2264 for maintaining out-of-tree patches for some time.
2266 PP/01 Bugzilla 139: Documentation and portability issues.
2267 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2268 Handle per-OS dynamic-module compilation flags.
2270 PP/02 Let /dev/null have normal permissions.
2271 The 4.73 fixes were a little too stringent and complained about the
2272 permissions on /dev/null. Exempt it from some checks.
2273 Reported by Andreas M. Kirchwitz.
2275 PP/03 Report version information for many libraries, including
2276 Exim version information for dynamically loaded libraries. Created
2277 version.h, now support a version extension string for distributors
2278 who patch heavily. Dynamic module ABI change.
2280 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2281 privilege escalation vulnerability whereby the Exim run-time user
2282 can cause root to append content of the attacker's choosing to
2285 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2288 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2289 If dropping privileges for untrusted macros, we disabled normal logging
2290 on the basis that it would fail; for the Exim run-time user, this is not
2291 the case, and it resulted in successful deliveries going unlogged.
2292 Fixed. Reported by Andreas Metzler.
2298 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2299 only prepend for the Resent-* case. Fixes regression introduced in
2300 Exim 4.70 by NM/22 for Bugzilla 607.
2302 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2303 increasing numbers of administrators be bitten by this.
2305 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2307 PP/03 Bugzilla 994: added openssl_options main configuration option.
2309 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2311 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2313 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2315 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2316 without a peer certificate, leading to a segfault because of an
2317 assumption that peers always have certificates. Be a little more
2318 paranoid. Problem reported by Martin Tscholak.
2320 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2321 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2322 NB: ClamAV planning to remove STREAM in "middle of 2010".
2323 CL also introduces -bmalware, various -d+acl logging additions and
2324 more caution in buffer sizes.
2326 PP/09 Implemented reverse_ip expansion operator.
2328 PP/10 Bugzilla 937: provide a "debug" ACL control.
2330 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2332 PP/12 Bugzilla 973: Implement --version.
2334 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2336 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2338 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2340 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2341 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2342 ignore trailing whitespace.
2344 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2346 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2349 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2350 Notification from Dr Andrew Aitchison.
2352 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2353 ExtendedDetectionInfo response format.
2354 Notification from John Horne.
2356 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2359 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2360 XSL and documented dependency on system catalogs, with examples of how
2363 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2366 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2367 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2368 configuration file which is writeable by the Exim user or group.
2370 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2371 of configuration files to cover files specified with the -C option if
2372 they are going to be used with root privileges, not just the default
2375 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2376 option (effectively making it always true).
2378 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2379 files to be used while preserving root privileges.
2381 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2382 that rogue child processes cannot use them.
2384 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2385 run-time user, instead of root.
2387 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2388 Exim run-time user without dropping privileges.
2390 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2391 result string, instead of calling string_vformat() twice with the same
2394 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2395 for other users. Others should always drop root privileges if they use
2396 -C on the command line, even for a whitelisted configure file.
2398 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2400 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2406 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2407 $data_path, and $header_path variables; fixed documentation bugs and
2410 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2411 exipick to access non-standard spools, including the "frozen" queue
2414 NM/01 Bugzilla 965: Support mysql stored procedures.
2415 Patch from Alain Williams
2417 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2419 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2420 Patch from Andreas Metzler
2422 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2423 Patch from Kirill Miazine
2425 NM/05 Bugzilla 671: Added umask to procmail example.
2427 JJ/03 installed exipick 20100323.0, fixing doc bug
2429 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2430 directory. Notification and patch from Dan Rosenberg.
2432 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2434 TK/02 Improve log output when DKIM signing operation fails.
2436 MH/01 Treat the transport option dkim_domain as a colon separated
2437 list, not as a single string, and sign the message with each element,
2438 omitting multiple occurences of the same signer.
2440 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2441 Bugzilla 985, 986. Patch by Simon Arlott
2443 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2444 Patch by Simon Arlott
2446 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2447 MBX locking. Notification from Dan Rosenberg.
2453 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2455 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2457 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2459 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2461 NM/04 Bugzilla 915: Flag broken perl installation during build.
2467 TK/01 Added patch by Johannes Berg that expands the main option
2468 "spamd_address" if it starts with a dollar sign.
2470 TK/02 Write list of recipients to X-Envelope-Sender header when building
2471 the mbox-format spool file for content scanning (suggested by Jakob
2474 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2475 (http://www.dcc-servers.net/) support via dccifd. Activated by
2476 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2478 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2479 by Mark Daniel Reidel <mr@df.eu>.
2481 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2482 When building exim an external PCRE library is now needed -
2483 PCRE is a system library on the majority of modern systems.
2484 See entry on PCRE_LIBS in EDITME file.
2486 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2487 conversation. Added nologin parameter to request.
2488 Patch contributed by Kirill Miazine.
2490 TF/01 Do not log submission mode rewrites if they do not change the address.
2492 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2494 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2495 log files in place. Contributed by Roberto Lima.
2497 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2499 TF/03 Bugzilla 615: When checking the local_parts router precondition
2500 after a local_part_suffix or local_part_prefix option, Exim now
2501 does not use the address's named list lookup cache, since this
2502 contains cached lookups for the whole local part.
2504 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2505 Robert Millan. Documentation is in experimental-spec.txt.
2507 TF/04 Bugzilla 668: Fix parallel build (make -j).
2509 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2511 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2512 Patch provided by Jan Srzednicki.
2514 TF/05 Leading white space used to be stripped from $spam_report which
2515 wrecked the formatting. Now it is preserved.
2517 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2518 that they are available at delivery time.
2520 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2522 TF/08 TLS error reporting now respects the incoming_interface and
2523 incoming_port log selectors.
2525 TF/09 Produce a more useful error message if an SMTP transport's hosts
2526 setting expands to an empty string.
2528 NM/06 Bugzilla 744: EXPN did not work under TLS.
2529 Patch provided by Phil Pennock.
2531 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2532 Patch provided by Richard Godbee.
2534 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2535 acl_smtp_notquit, added index entry.
2537 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2538 Patch provided by Eugene Bujak.
2540 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2541 accept(). Patch provided by Maxim Dounin.
2543 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2544 Patch provided by Phil Pennock.
2546 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2548 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2549 Patch provided by Brad "anomie" Jorsch.
2551 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2552 Patch provided by Dean Brooks.
2554 TK/05 Add native DKIM support (does not depend on external libraries).
2556 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2557 Patch provided by Graeme Fowler.
2559 NM/16 Bugzilla 851: Documentation example syntax fix.
2561 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2563 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2566 NM/19 Bugzilla 745: TLS version reporting.
2567 Patch provided by Phil Pennock.
2569 NM/20 Bugzilla 167: bool: condition support.
2570 Patch provided by Phil Pennock.
2572 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2573 clients. Patch provided by Phil Pennock.
2575 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2576 Patch provided by Brad "anomie" Jorsch.
2578 NM/23 Bugzilla 687: Fix misparses in eximstats.
2579 Patch provided by Heiko Schlittermann.
2581 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2582 Patch provided by Heiko Schlittermann.
2584 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2585 plus update to original patch.
2587 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2589 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2590 Patch provided by David Brownlee.
2592 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2594 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2596 NM/30 Bugzilla 888: TLS documentation bugfixes.
2598 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2600 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2601 Unlike the original bugzilla I have changed all shell scripts in src tree.
2603 NM/33 Bugzilla 898: Transport filter timeout fix.
2604 Patch by Todd Rinaldo.
2606 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2607 Patch by Serge Demonchaux.
2609 NM/35 Bugzilla 39: Base64 decode bug fixes.
2610 Patch by Jakob Hirsch.
2612 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2614 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2616 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2618 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2624 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2625 ALT-N's libdkim that I have put here:
2626 http://duncanthrax.net/exim-experimental/
2628 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2629 were called 'true' and it seems that C99 defines that as a reserved
2630 keyword to be used with 'bool' variable types. That means you could
2631 not include C99-style headers which use bools without triggering
2632 build errors in sieve.c.
2634 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2635 as mailq or other aliases. Changed the --help handling significantly
2636 to do whats expected. exim_usage() emits usage/help information.
2638 SC/01 Added the -bylocaldomain option to eximstats.
2640 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2642 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2644 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2650 PH/01 Another patch from the Sieve maintainer.
2652 PH/02 When an IPv6 address is converted to a string for single-key lookup
2653 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2654 dots are used instead of colons so that keys in lsearch files need not
2655 contain colons. This was done some time before quoting was made available
2656 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2657 (notated using the quote facility) so as to distinguish them from IPv4
2658 keys. This meant that lookups for IP addresses in host lists did not work
2659 for iplsearch lookups.
2661 This has been fixed by arranging for IPv6 addresses to be expressed with
2662 colons if the lookup type is iplsearch. This is not incompatible, because
2663 previously such lookups could never work.
2665 The situation is now rather anomalous, since one *can* have colons in
2666 ordinary lsearch keys. However, making the change in all cases is
2667 incompatible and would probably break a number of configurations.
2669 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2672 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2673 conversion specification without a maximum field width, thereby enabling
2674 a rogue spamd server to cause a buffer overflow. While nobody in their
2675 right mind would setup Exim to query an untrusted spamd server, an
2676 attacker that gains access to a server running spamd could potentially
2677 exploit this vulnerability to run arbitrary code as the Exim user.
2679 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2680 $primary_hostname instead of what libspf2 thinks the hosts name is.
2682 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2683 a directory entry by the name of the lookup key. Previously, if a
2684 symlink pointed to a non-existing file or a file in a directory that
2685 Exim lacked permissions to read, a lookup for a key matching that
2686 symlink would fail. Now it is enough that a matching directory entry
2687 exists, symlink or not. (Bugzilla 503.)
2689 PH/03 The body_linecount and body_zerocount variables are now exported in the
2692 PH/04 Added the $dnslist_matched variable.
2694 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2695 This means they are set thereafter only if the connection becomes
2698 PH/06 Added the client_condition to authenticators so that some can be skipped
2699 by clients under certain conditions.
2701 PH/07 The error message for a badly-placed control=no_multiline_responses left
2702 "_responses" off the end of the name.
2704 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2706 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2707 (without spaces) instead of just copying the configuration text.
2709 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2711 PH/11 Added $max_received_linelength.
2713 PH/12 Added +ignore_defer and +include_defer to host lists.
2715 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2716 way in which PCRE > 7.0 is built.
2718 PH/14 Implemented queue_only_load_latch.
2720 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2721 MAIL command. The effect was to mangle the value on 64-bit systems.
2723 PH/16 Another patch from the Sieve maintainer.
2725 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2727 PH/18 If a system quota error occurred while trying to create the file for
2728 a maildir delivery, the message "Mailbox is full" was not appended to the
2729 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2730 only to a quota excession during the actual writing of the file.
2732 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2733 characters?) which causes problems in log lines. The DN values are now
2734 passed through string_printing() before being added to log lines.
2736 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2737 and InterBase are left for another time.)
2739 PH/21 Added message_body_newlines option.
2741 PH/22 Guard against possible overflow in moan_check_errorcopy().
2743 PH/23 POSIX allows open() to be a macro; guard against that.
2745 PH/24 If the recipient of an error message contained an @ in the local part
2746 (suitably quoted, of course), incorrect values were put in $domain and
2747 $local_part during the evaluation of errors_copy.
2753 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2754 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2757 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2758 issue a MAIL command.
2760 PH/02 In an ACL statement such as
2762 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2764 if a client was not listed at all, or was listed with a value other than
2765 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2766 the condition was not true (as it should be), so access was not denied.
2767 The bug was that the ! inversion was incorrectly passed on to the second
2768 item. This has been fixed.
2770 PH/03 Added additional dnslists conditions == and =& which are different from
2771 = and & when the dns lookup returns more than one IP address.
2773 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2774 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2776 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2777 FSYNC, which compiles an option called disable_fsync that allows for
2778 bypassing fsync(). The documentation is heavily laced with warnings.
2780 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2782 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2783 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2784 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2785 including adding "make clean"; (3) Added -fPIC when compiling the test
2786 dynamically loaded module, to get rid of a warning.
2788 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2789 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2790 = 0s. The bug is otherwise harmless.
2792 PH/07 There was a bug in the dovecot authenticator such that the value of
2793 $auth1 could be overwritten, and so not correctly preserved, after a
2794 successful authentication. This usually meant that the value preserved by
2795 the server_setid option was incorrect.
2797 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2799 PH/09 Installed PCRE release 7.0.
2801 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2802 run for batched SMTP input. It is now run at the start of every message
2803 in the batch. While fixing this I discovered that the process information
2804 (output by running exiwhat) was not always getting set for -bs and -bS
2805 input. This is fixed, and it now also says "batched" for BSMTP.
2807 PH/11 Added control=no_pipelining.
2809 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2810 patch, slightly modified), and move the expansion of helo_data till after
2811 the connection is made in the smtp transport (so it can use these
2814 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2816 PH/14 Added log_selector = +pid.
2818 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2820 PH/16 Add ${if forany and ${if forall.
2822 PH/17 Added dsn_from option to vary the From: line in DSNs.
2824 PH/18 Flush SMTP output before performing a callout, unless control =
2825 no_callout_flush is set.
2827 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2828 was true (the default) a successful delivery failed to delete the retry
2829 item, thus causing premature timeout of the address. The bug is now
2832 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2834 PH/21 Long custom messages for fakedefer and fakereject are now split up
2835 into multiline responses in the same way that messages for "deny" and
2836 other ACL rejections are.
2838 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2839 with slight modification.
2841 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2842 draft, changing the syntax and factoring some duplicate code".
2844 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2845 for deliveries of the second and subsequent messages over the same SMTP
2848 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2849 ${reduce, with only minor "tidies".
2851 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2853 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2854 expansion side effects.
2856 PH/27 When a message times out after an over-quota error from an Exim-imposed
2857 quota, the bounce message says "mailbox is full". This message was not
2858 being given when it was a system quota that was exceeded. It now should
2861 MH/03 Made $recipients available in local_scan(). local_scan() already has
2862 better access to the recipient list through recipients_list[], but
2863 $recipients can be useful in postmaster-provided expansion strings.
2865 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2866 in the case of a MAIL command with additional options following the
2867 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2868 were accidentally chopped off.
2870 PH/29 SMTP synchronization checks are implemented when a command is read -
2871 there is a check that no more input is waiting when there shouldn't be
2872 any. However, for some commands, a delay in an ACL can mean that it is
2873 some time before the response is written. In this time, more input might
2874 arrive, invalidly. So now there are extra checks after an ACL has run for
2875 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2876 pipelining has not been advertised.
2878 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2880 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2881 returned in the SMTP response when smtp_return_error_details was false.
2882 This has been fixed.
2884 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2885 instead of the C I/O that was originally supplied, because problems were
2886 reported on Solaris.
2888 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2889 Exim which did not show up earlier: it was assuming that a call to
2890 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2891 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2892 was a macro that became an assignment, so it seemed to work. This has
2893 changed to a proper function call with a void return, hence the compile
2894 error. Exim's code has been fixed.
2896 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2899 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2901 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2903 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2904 because it turns out to be too big or there is a timeout) there is no
2905 "Completed" line in the log. When some messages of this type were
2906 selected by exigrep, they were listed as "not completed". Others were
2907 picked up by some special patterns. I have improved the selection
2908 criteria to be more general.
2910 PH/38 The host_find_failed option in the manualroute router can now be set
2911 to "ignore", to completely ignore a host whose IP address cannot be
2912 found. If all hosts are ignored, the behaviour is controlled by the new
2913 host_all_ignored option.
2915 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2916 homing or because of multiple MX records with /mx) generated more than
2917 one IP address, and the following item turned out to be the local host,
2918 all the secondary addresses of the first item were incorrectly removed
2919 from the list, along with the local host and any following hosts (which
2920 is what is supposed to happen).
2922 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2923 whoever called Exim into the -H file. In the case of the daemon it was
2924 behaving confusingly. When first started, it used values for whoever
2925 started the daemon, but after a SIGHUP it used the Exim user (because it
2926 calls itself on a restart). I have changed the code so that it now always
2929 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2930 message are rejected with the same error (e.g. no authentication or bad
2931 sender address), and a DATA command is nevertheless sent (as can happen
2932 with PIPELINING or a stupid MUA), the error message that was given to the
2933 RCPT commands is included in the rejection of the DATA command. This is
2934 intended to be helpful for MUAs that show only the final error to their
2937 PH/42 Another patch from the Sieve maintainer.
2939 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2940 Eximstats - Fixed some broken HTML links and added missing column headers
2942 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2943 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2945 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2947 PH/43 Yet another patch from the Sieve maintainer.
2949 PH/44 I found a way to check for a TCP/IP connection going away before sending
2950 the response to the final '.' that terminates a message, but only in the
2951 case where the client has not sent further data following the '.'
2952 (unfortunately, this is allowed). However, in many cases there won't be
2953 any further data because there won't be any more messages to send. A call
2954 to select() can be used: if it shows that the input is "ready", there is
2955 either input waiting, or the socket has been closed. An attempt to read
2956 the next input character can distinguish the two cases. Previously, Exim
2957 would have sent an OK response which the client would never have see.
2958 This could lead to message repetition. This fix should cure that, at
2959 least in a lot of common cases.
2961 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2962 advertised in response to EHLO.
2968 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2969 fixed by 4.65/MH/01 (is this a record?) are fixed:
2971 (i) An empty string was always treated as zero by the numeric comparison
2972 operators. This behaviour has been restored.
2974 (ii) It is documented that the numeric comparison operators always treat
2975 their arguments as decimal numbers. This was broken in that numbers
2976 starting with 0 were being interpreted as octal.
2978 While fixing these problems I realized that there was another issue that
2979 hadn't been noticed. Values of message_size_limit (both the global option
2980 and the transport option) were treated as octal if they started with 0.
2981 The documentation was vague. These values are now always treated as
2982 decimal, and I will make that clear in the documentation.
2988 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2989 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2992 MH/01 Don't check that the operands of numeric comparison operators are
2993 integers when their expansion is in "skipping" mode (fixes bug
2994 introduced by 4.64-PH/07).
2996 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2997 child addresses, Exim now panics and dies. Previously, because the count
2998 is held in a short int, deliveries were likely to be lost. As such a
2999 large number of recipients for a single message is ridiculous
3000 (performance will be very, very poor), I have chosen to impose a limit
3001 rather than extend the field.
3007 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3008 leftover -K file (the existence of which was triggered by #402).
3009 While we were at it, introduced process PID as part of the -K
3010 filename. This should rule out race conditions when creating
3013 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3014 processing considerably. Previous code took too long for large mails,
3015 triggering a timeout which in turn triggers #401.
3017 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3018 in the DK code in transports.c. sendfile() is not really portable,
3019 hence the _LINUX specificness.
3021 TF/01 In the add_headers option to the mail command in an Exim filter,
3022 there was a bug that Exim would claim a syntax error in any
3023 header after the first one which had an odd number of characters
3026 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3027 callout verification, Exim cached a "reject" for the entire domain. This
3028 is correct for most verifications, but it is not correct for a recipient
3029 verification with use_sender or use_postmaster set, because in that case
3030 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3031 case of MAIL FROM:<> rejection from other early rejections (e.g.
3032 rejection of HELO). When verifying a recipient using a non-null MAIL
3033 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3034 Whatever the result of the callout, the value of the domain cache is
3035 left unchanged (for any other kind of callout, getting as far as trying
3036 RCPT means that the domain itself is ok).
3038 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3041 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3042 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3043 session, the daemon ignores ECONNECT errors and logs others; it now
3044 ignores EPIPE as well.
3046 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3047 (quoted-printable decoding).
3049 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3050 later the small subsequent patch to fix an introduced bug.
3052 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3054 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3056 PH/08 An error is now given if message_size_limit is specified negative.
3058 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3059 to be given (somewhat) arbitrary names.
3061 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3064 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3065 miscellaneous code fixes
3067 PH/10 Added the log_reject_target ACL modifier to specify where to log
3070 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3071 hostname. This is wrong, because it relates to the incoming message (and
3072 probably the interface on which it is arriving) and not to the outgoing
3073 callout (which could be using a different interface). This has been
3074 changed to use the value of the helo_data option from the smtp transport
3075 instead - this is what is used when a message is actually being sent. If
3076 there is no remote transport (possible with a router that sets up host
3077 addresses), $smtp_active_hostname is used.
3079 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3080 tweaks were necessary in order to get it to work (see also 21 below):
3081 (a) The code assumed that strncpy() returns a negative number on buffer
3082 overflow, which isn't the case. Replaced with Exim's string_format()
3084 (b) There were several signed/unsigned issues. I just did the minimum
3085 hacking in of casts. There is scope for a larger refactoring.
3086 (c) The code used strcasecmp() which is not a standard C function.
3087 Replaced with Exim's strcmpic() function.
3088 (d) The code set only $1; it now sets $auth1 as well.
3089 (e) A simple test gave the error "authentication client didn't specify
3090 service in request". It would seem that Dovecot has changed its
3091 interface. Fortunately there's a specification; I followed it and
3092 changed what the client sends and it appears to be working now.
3094 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3097 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3098 address A is aliased to B and C, where B exists and C does not. Without
3099 -v the output is "A verified" because verification stops after a
3100 successful redirection if more than one address is generated. However,
3101 with -v the child addresses are also verified. Exim was outputting "A
3102 failed to verify" and then showing the successful verification for C,
3103 with its parentage. It now outputs "B failed to verify", showing B's
3104 parentage before showing the successful verification of C.
3106 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3107 look up a TXT record in a specific list after matching in a combined
3110 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3111 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3112 they consult the DNS. I had assumed they would set it the way they
3113 wanted; and indeed my experiments on Linux seem to show that in some
3114 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3115 To be on the safe side, however, I have now made the interface to
3116 host_find_byname() similar to host_find_bydns(), with an argument
3117 containing the DNS resolver options. The host_find_byname() function now
3118 sets these options at its start, just as host_find_bydns() does. The smtp
3119 transport options dns_qualify_single and dns_search_parents are passed to
3120 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3121 of host_find_byname() use the default settings of RES_DEFNAMES
3122 (qualify_single) but not RES_DNSRCH (search_parents).
3124 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3125 spool_read_header() do less string testing, by means of a preliminary
3126 switch on the second character of optional "-foo" lines. (This is
3127 overdue, caused by the large number of possibilities that now exist.
3128 Originally there were few.) While I was there, I also converted the
3129 str(n)cmp tests so they don't re-test the leading "-" and the first
3130 character, in the hope this might squeeze out yet more improvement.
3132 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3133 flag allowing group syntax was set by the header_syntax check but not
3134 turned off, possible causing trouble later; (2) The flag was not being
3135 set at all for the header_verify test, causing "group"-style headers to
3136 be rejected. I have now set it in this case, and also caused header_
3137 verify to ignore an empty address taken from a group. While doing this, I
3138 came across some other cases where the code for allowing group syntax
3139 while scanning a header line wasn't quite right (mostly, not resetting
3140 the flag correctly in the right place). These bugs could have caused
3141 trouble for malformed header lines. I hope it is now all correct.
3143 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3144 with the "reply" argument non-NULL. The code, however (which originally
3145 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3146 but it didn't always do it. This confused somebody who was copying the
3147 code for some other use. I have removed all the tests.
3149 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3150 feature that was used to support insecure browsers during the U.S. crypto
3151 embargo. It requires special client support, and Exim is probably the
3152 only MTA that supported it -- and would never use it because real RSA is
3153 always available. This code has been removed, because it had the bad
3154 effect of slowing Exim down by computing (never used) parameters for the
3155 RSA_EXPORT functionality.
3157 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3158 authenticator to fail if there's a tab character in the incoming data
3159 (there should never be unless someone is messing about, as it's supposed
3160 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3161 if the connection is using TLS or if the remote IP is the same as the
3162 local IP, and the "valid-client-cert option" if a client certificate has
3165 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3166 authenticators. This can be used for authorization after authentication
3167 succeeds. (In the case of plaintext, it servers for both authentication
3170 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3171 if any retry times were supplied.
3173 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3174 connection, where there is no client IP address to check. In this
3175 situation, the verify now always succeeds.
3177 PH/25 Applied John Jetmore's -Mset patch.
3179 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3181 PH/27 In a string expansion for a processed (not raw) header when multiple
3182 headers of the same name were present, leading whitespace was being
3183 removed from all of them, but trailing whitespace was being removed only
3184 from the last one. Now trailing whitespace is removed from each header
3185 before concatenation. Completely empty headers in a concatenation (as
3186 before) are ignored.
3188 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3189 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3191 PH/29 [Removed. This was a change that I later backed out, and forgot to
3192 correct the ChangeLog entry (that I had efficiently created) before
3193 committing the later change.]
3195 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3196 address errors (4xx response to RCPT) over the same connection as other
3197 messages routed to the same hosts. Such deliveries are always "forced",
3198 so retry times are not inspected. This resulted in far too many retries
3199 for the affected addresses. The effect occurred only when there were more
3200 hosts than the hosts_max_try setting in the smtp transport when it had
3201 the 4xx errors. Those hosts that it had tried were not added to the list
3202 of hosts for which the message was waiting, so if all were tried, there
3203 was no problem. Two fixes have been applied:
3205 (i) If there are any address or message errors in an SMTP delivery, none
3206 of the hosts (tried or untried) are now added to the list of hosts
3207 for which the message is waiting, so the message should not be a
3208 candidate for sending over the same connection that was used for a
3209 successful delivery of some other message. This seems entirely
3210 reasonable: after all the message is NOT "waiting for some host".
3211 This is so "obvious" that I'm not sure why it wasn't done
3212 previously. Hope I haven't missed anything, but it can't do any
3213 harm, as the worst effect is to miss an optimization.
3215 (ii) If, despite (i), such a delivery is accidentally attempted, the
3216 routing retry time is respected, so at least it doesn't keep
3217 hammering the server.
3219 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3220 in ${readsocket because some servers need this prod.
3222 PH/32 Added some extra debug output when updating a wait-xxx database.
3224 PH/33 The hint "could be header name not terminated by colon", which has been
3225 given for certain expansion errors for a long time, was not being given
3226 for the ${if def:h_colon_omitted{... case.
3228 PH/34 The spec says: "With one important exception, whenever a domain list is
3229 being scanned, $domain contains the subject domain." There was at least
3230 one case where this was not true.
3232 PH/35 The error "getsockname() failed: connection reset by peer" was being
3233 written to the panic log as well as the main log, but it isn't really
3234 panic-worthy as it just means the connection died rather early on. I have
3235 removed the panic log writing for the ECONNRESET error when getsockname()
3238 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3239 runs only) independently of the message's sender address. This meant
3240 that, if the 4xx error was in fact related to the sender, a different
3241 message to the same recipient with a different sender could confuse
3242 things. In particular, this can happen when sending to a greylisting
3243 server, but other circumstances could also provoke similar problems.
3244 I have changed the default so that the retry time for these errors is now
3245 based a combination of the sender and recipient addresses. This change
3246 can be overridden by setting address_retry_include_sender=false in the
3249 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3250 remote server are returned as part of bounce messages. This was not
3251 happening for LMTP over a pipe (the lmtp transport), but now it is the
3252 same for both kinds of LMTP.
3254 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3255 in header lines that were in fact CNAMEs. This is no longer the case.
3257 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3258 and queue runs started by the daemon processed all messages. This has
3259 been fixed so that -R and -S can now usefully be given with -q<time>.
3261 PH/40 Import PCRE release 6.7 (fixes some bugs).
3263 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3265 PH/42 Give an error if -q is specified more than once.
3267 PH/43 Renamed the variables $interface_address and $interface_port as
3268 $received_ip_address and $received_port, to make it clear that these
3269 values apply to message reception, and not to the outgoing interface when
3270 a message is delivered. (The old names remain recognized, of course.)
3272 PH/44 There was no timeout on the connect() call when using a Unix domain
3273 socket in the ${readsocket expansion. There now is.
3275 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3276 be meaningful with "accept".
3278 SC/01 Eximstats V1.43
3279 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3281 SC/02 Eximstats V1.44
3282 Use a glob alias rather than an array ref in the generated
3283 parser. This improves both readability and performance.
3285 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3286 Collect SpamAssassin and rejection statistics.
3287 Don't display local sender or destination tables unless
3288 there is data to show.
3289 Added average volumes into the top table text output.
3291 SC/04 Eximstats V1.46
3292 Collect data on the number of addresses (recipients)
3293 as well as the number of messages.
3295 SC/05 Eximstats V1.47
3296 Added 'Message too big' to the list of mail rejection
3297 reasons (thanks to Marco Gaiarin).
3299 SC/06 Eximstats V1.48
3300 Mainlog lines which have GMT offsets and are too short to
3301 have a flag are now skipped.
3303 SC/07 Eximstats V1.49 (Alain Williams)
3304 Added the -emptyok flag.
3306 SC/08 Eximstats V1.50
3307 Fixes for obtaining the IP address from reject messages.
3309 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3310 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3311 whitespace changes from 4.64-PH/27
3313 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3316 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3317 are found, allow negative numbers in numeric criteria)
3319 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3321 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3322 to match changes made in 4.64-PH/43
3324 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3326 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3327 "domain" is actually an IP address, and force a failure. This locks out
3328 those revolvers/nameservers that support "A-for-A" lookups, in
3329 contravention of the specifications.
3331 PH/48 When a host name was looked up from an IP address, and the subsequent
3332 forward lookup of the name timed out, the host name was left in
3333 $sender_host_name, contrary to the specification.
3335 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3336 restricted to single-key lookups, Exim was not diagnosing an error if
3337 * or *@ was used with a query-style lookup.
3339 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3341 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3342 long ago, but noone interested enough thought of it. Let's just say that
3343 the "1.1" means that there are some new functions that weren't there at
3344 some point in the past.
3346 PH/51 Error processing for expansion failure of helo_data from an smtp
3347 transport during callout processing was broken.
3349 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3350 tested/used via the -bh/-bhc/-bs options.
3352 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3353 bug, fixed in subsequent PCRE releases).
3355 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3356 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3358 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3364 SC/01 Use a glob alias rather than an array ref in eximstats generated
3365 parser. This improves both readability and performance.
3367 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3368 Don't display local sender or destination tables in eximstats unless
3369 there is data to show.
3370 Added average volumes into the eximstats top table text output.
3372 SC/03 Collect data on the number of addresses (recipients) as well
3373 as the number of messages in eximstats.
3375 TF/01 Correct an error in the documentation for the redirect router. Exim
3376 does (usually) call initgroups() when daemonizing.
3378 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3379 with consistent privilege compared to when running as a daemon.
3381 TF/03 Note in the spec that $authenticated_id is not set for local
3382 submissions from trusted users.
3384 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3385 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3387 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3388 by adding some example configuration directives to the default
3389 configuration file. A little bit of work is required to uncomment the
3390 directives and define how usernames and passwords are checked, but
3391 there is now a framework to start from.
3393 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3394 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3395 without this. I don't know how relevant this is to other LDAP libraries.
3397 PH/02 Add the verb name to the "unknown ACL verb" error.
3399 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3401 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3403 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3404 directory not expanded when it should be if an expanded home directory
3405 was set for the address (which is overridden by the transport).
3407 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3410 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3411 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3412 because it is too late at that time, and has no effect.
3414 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3415 security issue with \' (bugzilla #107). I could not use the
3416 PQescapeStringConn() function, because it needs a PGconn value as one of
3419 PH/08 When testing addresses using -bt, indicate those final addresses that
3420 are duplicates that would not cause an additional delivery. At least one
3421 person was confused, thinking that -bt output corresponded to deliveries.
3422 (Suppressing duplicates isn't a good idea as you lose the information
3423 about possibly different redirections that led to the duplicates.)
3425 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3426 systems where poll() doesn't work, in particular OS X.
3428 PH/10 Added more information to debugging output for retry time not reached.
3430 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3431 operations in malware.c.
3433 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3436 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3437 syslog_duplication set false, log lines that would normally be written
3438 both the the main log and to the reject log were not written to syslog at
3441 PH/14 In the default configuration, change the use of "message" in ACL warn
3442 statements to "add_header".
3444 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3445 not followed by a command (e.g. "seen endif").
3447 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3448 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3451 PH/17 Added extra conditions to the default value of delay_warning_condition
3455 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3456 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3457 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3460 The Auto-Submitted: and various List- headers are standardised, whereas I
3461 don't think Precedence: ever was.
3463 PH/18 Refactored debugging code in route_finduser() to show more information,
3464 in particular, the error code if getpwnam() issues one.
3466 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3467 This is apparently needed in addition to the PH/07 change above to avoid
3468 any possible encoding problems.
3470 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3471 but not after initializing Perl.
3473 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3474 output them only if debugging. By default they are written stderr,
3475 apparently, which is not desirable.
3477 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3480 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3483 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3485 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3486 authenticated or an ident call has been made. Suppress the default
3487 values for $authenticated_id and $authenticated_sender (but permit -oMai
3488 and -oMas) when testing with -bh.
3490 PH/24 Re-jigged the order of the tests in the default configuration so that the
3491 tests for valid domains and recipients precede the DNS black list and CSA
3492 tests, on the grounds that those ones are more expensive.
3494 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3495 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3496 This bug exists in every version of Exim that I still have, right back to
3499 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3500 However, an attempt to turn on case-sensitivity in a regex key by
3501 including (?-i) didn't work because the subject string was already
3502 lowercased, and the effects were non-intuitive. It turns out that a
3503 one-line patch can be used to allow (?-i) to work as expected.
3509 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3510 other effects) broke the use of negated acl sub-conditions.
3512 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3515 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3516 "Deny" causes Exim to reject the incoming connection with a 554 error.
3517 Unfortunately, if there is a major crisis, such as a disk failure,
3518 tcp-wrappers gives "deny", whereas what one would like would be some
3519 kind of temporary error. A kludge has been added to help with this.
3520 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3521 554 error is used if errno is still zero or contains ENOENT (which occurs
3522 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3525 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3527 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3528 errors. Otherwise a message that provokes a temporary error (when other
3529 messages do not) can cause a whole host to time out.
3531 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3532 addresses were routed directly to files or pipes from a redirect router.
3533 File deliveries just didn't batch; pipe deliveries might have suffered
3536 PH/06 A failure to get a lock for a hints database would erroneously always say
3537 "Failed to get write lock", even when it was really a read lock.
3539 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3540 of 0600. This has been changed to use the value of the lockfile_mode
3541 option (which defaults to 0600).
3543 PH/08 Applied small patch from the Sieve maintainer.
3545 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3546 folder from quota calculations, a direct delivery into this folder messed
3547 up the contents of the maildirsize file. This was because the regex was
3548 used only to exclude .Trash (or whatever) when the size of the mailbox
3549 was calculated. There was no check that a delivery was happening into an
3550 excluded directory. This bug has been fixed by ignoring all quota
3551 processing for deliveries into excluded directories.
3553 PH/10 Added the maildirfolder_create_regex option to appendfile.
3559 PH/01 The code for finding all the local interface addresses on a FreeBSD
3560 system running IPv6 was broken. This may well have applied to all BSD
3561 systems, as well as to others that have similar system calls. The broken
3562 code found IPv4 interfaces correctly, but gave incorrect values for the
3563 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3564 that it would not match correctly against @[] and not recognize the IPv6
3567 PH/02 The ipliteral router was not recognizing addresses of the form user@
3568 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3570 PH/03 Added disable_ipv6.
3572 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3573 decoded form, because it is most often used to construct To: headers
3574 lines in autoreplies, and the decoded form may well be syntactically
3575 invalid. However, $reply_address has leading white space removed, and all
3576 newlines turned into spaces so that the autoreply transport does not
3579 PH/05 If group was specified without a user on a router, and no group or user
3580 was specified on a transport, the group from the router was ignored.
3582 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3583 for visible compile-time settings that can be used to change these
3584 numbers, for those that want even more. Backwards compatibility with old
3585 spool files has been maintained. However, going back to a previous Exim
3586 release will lost any variables that are in spool files.
3588 PH/07 Two small changes when running in the test harness: increase delay when
3589 passing a TCP/IP connection to a new process, in case the original
3590 process has to generate a bounce, and remove special handling of
3591 127.0.0.2 (sic), which is no longer necessary.
3593 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3594 be the same on different OS.
3596 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3599 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3600 whether --show-vars was specified or not
3602 JJ/02 exipick: Added support for new ACL variable spool format introduced
3605 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3606 syntactically invalid From: or Reply-to: line, and a filter used this to
3607 generate an autoreply, and therefore failed to obtain an address for the
3608 autoreply, Exim could try to deliver to a non-existent relative file
3609 name, causing unrelated and misleading errors. What now happens is that
3610 it logs this as a hard delivery error, but does not attempt to create a
3613 PH/11 The exinext utility has a -C option for testing purposes, but although
3614 the given file was scanned by exinext itself; it wasn't being passed on
3615 when Exim was called.
3617 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3618 an end-of-file indication when reading a command response.
3620 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3621 compiled. In many other places in Exim, IPv6 addresses are always
3622 recognized, so I have changed this. It also means that IPv4 domain
3623 literals of the form [IPV4:n.n.n.n] are now always recognized.
3625 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3626 used if the router is not running as root, for example, when verifying at
3627 ACL time, or when using -bh. The debugging output from this situation was
3628 non-existent - all you got was a failure to exec. I have made two
3631 (a) Failures to set uid/gid, the current directory, or a process leader
3632 in a subprocess such as that created by queryprogram now generate
3633 suitable debugging output when -d is set.
3635 (b) The queryprogram router detects when it is not running as root,
3636 outputs suitable debugging information if -d is set, and then runs
3637 the subprocess without attempting to change uid/gid.
3639 PH/15 Minor change to Makefile for building test_host (undocumented testing
3642 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3643 additional section of a DNS packet that returns MX or SRV records.
3644 Instead, it always explicitly searches for A/AAAA records. This avoids
3645 major problems that occur when a DNS server includes only records of one
3646 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3647 fixed another bug: if SRV records were looked up and the corresponding
3648 address records were *not* found in the additional section, the port
3649 values from the SRV records were lost.
3651 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3652 using the correct key (the original address) when searching the retry
3653 rules in order to find which one to use for generating the retry hint.
3655 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3656 adding the default one. Similarly, if it contains a Reply-To: header, the
3657 errors_reply_to option, if set, is not used.
3659 PH/19 When calculating a retry time, Exim used to measure the "time since
3660 failure" by looking at the "first failed" field in the retry record. Now
3661 it does not use this if it is later than than the arrival time of the
3662 message. Instead it uses the arrival time. This makes for better
3663 behaviour in cases where some deliveries succeed, thus re-setting the
3664 "first failed" field. An example is a quota failure for a huge message
3665 when small messages continue to be delivered. Without this change, the
3666 "time since failure" will always be short, possible causing more frequent
3667 delivery attempts for the huge message than are intended.
3668 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3670 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3671 $1, $2, $3) because the numerical variables can be reset during some
3672 expansion items (e.g. "match"), thereby losing the authentication data.
3674 PH/21 Make -bV show the size of off_t variables so that the test suite can
3675 decide whether to run tests for quotas > 2G.
3677 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3678 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3679 filecount value is greater than 2G or if a quota value is greater than 2G
3680 on a system where the size of off_t is not greater than 4, a panic error
3683 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3684 never match. The debug and -bh output now contains an explicit error
3685 message indicating a malformed IPv4 address or mask.
3687 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3688 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3689 PH/23 above applies.
3691 PH/25 Do not write to syslog when running in the test harness. The only
3692 occasion when this arises is a failure to open the main or panic logs
3693 (for which there is an explicit test).
3695 PH/26 Added the /no_tell option to "control=freeze".
3697 PH/27 If a host name lookup failed very early in a connection, for example, if
3698 the IP address matched host_lookup and the reverse lookup yielded a name
3699 that did not have a forward lookup, an error message of the form "no IP
3700 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3701 could be logged. Now it outputs the IP address instead of "NULL".
3703 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3704 allows the sender and the authenticated sender to be set when
3705 submitting a message from within Exim. Since child_open_exim() is
3706 documented for local_scan(), the new function should be too.
3708 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3709 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3710 results in an empty string is now treated as unset.
3712 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3714 PH/31 Added sender_verify_fail logging option.
3716 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3717 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3718 tidied the source and removed it altogether.
3720 PH/33 When a queue run was abandoned because the load average was too high, a
3721 log line was always written; now it is written only if the queue_run log
3722 selector is set. In addition, the log line for abandonment now contains
3723 information about the queue run such as the pid. This is always present
3724 in "start" and "stop" lines but was omitted from the "abandon" line.
3726 PH/34 Omit spaces between a header name and the colon in the error message that
3727 is given when verify = headers_syntax fails (if there are lots of them,
3728 the message gets confusing).
3730 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3731 names, as there are now some PTR records that contain slashes. This check
3732 is only to protect against broken name servers that fall over on strange
3733 characters, so the fact that it applies to all lookups doesn't matter.
3735 PH/36 Now that the new test suite is complete, we can remove some of the
3736 special code in Exim that was needed for the old test suite. For example,
3737 sorting DNS records because real resolvers return them in an arbitrary
3738 order. The new test suite's fake resolver always returns records in the
3741 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3742 bounces) except when queue_only is set, to avoid logging races between
3743 the different processes.
3745 PH/38 Panic-die if .include specifies a non-absolute path.
3747 PH/39 A tweak to the "H" retry rule from its user.
3749 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3750 a label. They prevented compilation on older perls.
3752 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3753 a warning to be raised on newish perls.
3755 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3756 on queue. Changes to match documented behaviour of showing count of
3757 messages matching specified criteria.
3759 PH/40 Changed the default ident timeout from 30s to 5s.
3761 PH/41 Added support for the use of login_cap features, on those BSD systems
3762 that have them, for controlling the resources used by pipe deliveries.
3764 PH/42 The content-scanning code uses fopen() to create files in which to put
3765 message data. Previously it was not paying any attention to the mode of
3766 the files. Exim runs with umask(0) because the rest of the code creates
3767 files with open(), and sets the required mode explicitly. Thus, these
3768 files were ending up world-writeable. This was not a big issue, because,
3769 being within the spool directory, they were not world-accessible. I have
3770 created a function called modefopen, which takes an additional mode
3771 argument. It sets umask(777), creates the file, chmods it to the required
3772 mode, then resets the umask. All the relevant calls to fopen() in the
3773 content scanning code have been changed to use this function.
3775 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3776 to 24 hours. This avoids potential overflow problems when processing G
3777 and H retry rules. I suspect nobody ever tinkers with this value.
3779 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3781 PH/45 When the plaintext authenticator is running as a client, the server's
3782 challenges are checked to ensure they are valid base64 strings. By
3783 default, the authentication attempt is cancelled if an invalid string is
3784 received. Setting client_ignore_invalid_base64 true ignores these errors.
3785 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3786 they are received. Thus, the responses can be made to depend on the
3787 challenges. If an invalid string is ignored, an empty string is placed in
3790 PH/46 Messages that are created by the autoreply transport now contains a
3791 References: header, in accordance with RFCs 2822 and 3834.
3793 PH/47 Added authenticated_sender_force to the smtp transport.
3795 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3797 PH/49 Installed latest patch from the Sieve maintainer.
3799 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3800 was also set, the appendfile transport was unnecessarily scanning a
3801 directory of message files (e.g. for maildir delivery) to find the count
3802 of files (along with the size), even though it did not need this
3803 information. It now does the scan only if it needs to find either the
3804 size of the count of files.
3806 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3808 PH/52 Two bugs concerned with error handling when the smtp transport is
3811 (i) Exim was not creating retry information for temporary errors given
3812 for individual recipients after the DATA command when the smtp transport
3813 was used in LMTP mode. This meant that they could be retried too
3814 frequently, and not timed out correctly.
3816 (ii) Exim was setting the flag that allows error details to be returned
3817 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3818 recipients that were returned after the DATA command.
3820 PH/53 This is related to PH/52, but is more general: for any failing address,
3821 when detailed error information was permitted to be returned to the
3822 sender, but the error was temporary, then after the final timeout, only
3823 "retry timeout exceeded" was returned. Now it returns the full error as
3824 well as "retry timeout exceeded".
3826 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3827 do this, and (what is worse) MTAs that accept it.
3829 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3830 will now be deprecated.
3832 PH/56 New os.c-cygwin from the Cygwin maintainer.
3834 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3835 formats (previously only available in exim formats via -bpr, -bpru,
3836 and -bpra. Now also available in native and exiqgrep formats)
3838 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3839 with very large, slow to parse queues
3841 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3843 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3845 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3846 responses to authentication challenges, though it was showing the
3847 challenges; (ii) I've removed the CR characters from the debug output for
3850 PH/58 Allow for the insertion of a newline as well as a space when a string
3851 is turned into more than one encoded-word during RFC 2047 encoding. The
3852 Sieve code now uses this.
3854 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3855 data_4xx, lost_connection, tls_required.
3857 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3858 message was being sent as an SMTP response.
3860 PH/61 Add -l and -k options to exicyclog.
3862 PH/62 When verifying, if an address was redirected to one new address, so that
3863 verification continued, and the new address failed or deferred after
3864 having set something in $address_data, the value of $address_data was not
3865 passed back to the ACL. This was different to the case when no
3866 redirection occurred. The value is now passed back in both cases.
3868 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3869 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3870 use login_cap.h, so on its own it isn't the distinguishing feature. The
3871 new name refers directly to the setclassresources() function.
3873 PH/65 Added configuration files for NetBSD3.
3875 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3877 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3878 is preferred over IPv4.
3880 PH/68 The bounce_return_message and bounce_return_body options were not being
3881 honoured for bounces generated during the reception of non-SMTP messages.
3882 In particular, this applied to messages rejected by the ACL. This bug has
3883 been fixed. However, if bounce_return_message is true and bounce_return_
3884 body is false, the headers that are returned for a non-SMTP message
3885 include only those that have been read before the error was detected.
3886 (In the case of an ACL rejection, they have all been read.)
3888 PH/69 The HTML version of the specification is now built in a directory called
3889 spec_html instead of spec.html, because the latter looks like a path with
3890 a MIME-type, and this confuses some software.
3892 PH/70 Catch two compiler warnings in sieve.c.
3894 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3895 function verify_get_ident() calls ip_connect() to connect a socket, but
3896 if the "connect()" function timed out, ip_connect() used to close the
3897 socket. However, verify_get_ident() also closes the socket later, and in
3898 between Exim writes to the log, which may get opened at this point. When
3899 the socket was closed in ip_connect(), the log could get the same file
3900 descriptor number as the socket. This naturally causes chaos. The fix is
3901 not to close the socket in ip_connect(); the socket should be closed by
3902 the function that creates it. There was only one place in the code where
3903 this was missing, in the iplookup router, which I don't think anybody now
3904 uses, but I've fixed it anyway.
3906 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3907 well as to direct DNS lookups. Otherwise the handling of names in host
3908 lists is inconsistent and therefore confusing.
3914 PH/01 Two changes to the default runtime configuration:
3916 (1) Move the checks for relay_from_hosts and authenticated clients from
3917 after to before the (commented out) DNS black list checks.
3919 (2) Add control=submission to the relay_from_hosts and authenticated
3920 clients checks, on the grounds that messages accepted by these
3921 statements are most likely to be submissions.
3923 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3925 (1) Generate an error if the third argument for the ${prvs expansion is
3928 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3931 (3) Reset the variables that are obtained from the first argument of
3932 ${prvscheck and used in the second argument before leaving the code,
3933 because their memory is reclaimed, so using them afterwards may do
3936 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3937 one (it's much easier than Tom thought :-).
3939 (5) Because of (4), we can now allow for the use of $prvscheck_result
3940 inside the third argument.
3942 PH/03 For some reason, the default setting of PATH when running a command from
3943 a pipe transport was just "/usr/bin". I have changed it to
3946 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3947 anything to be listed in the output from -bV.
3949 PH/05 When a filter generated an autoreply, the entire To: header line was
3950 quoted in the delivery log line, like this:
3952 => >A.N.Other <ano@some.domain> <original@ddress> ...
3954 This has been changed so that it extracts the operative address. There
3955 may be more than one such address. If so, they are comma-separated, like
3958 => >ano@some.domain,ona@other.domain <original@ddress> ...
3960 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3961 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3962 not being looked up in the rDNS to get a host name, Exim was showing the
3963 IP address twice in Received: lines, even though the IP addresses were
3964 identical. For example:
3966 Received: from [1.2.3.4] (helo=[1.2.3.4])
3968 However, if the real host name was known, it was omitting the HELO data
3969 if it matched the actual IP address. This has been tidied up so that it
3970 doesn't show the same IP address twice.
3972 PH/07 When both +timestamp and +memory debugging was on, the value given by
3973 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3974 called by the debug printing, thereby overwriting the timestamp buffer.
3975 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3977 PH/08 When the original message was included in an autoreply transport, it
3978 always said "this is a copy of the message, including all the headers",
3979 even if body_only or headers_only was set. It now gives an appropriate
3982 PH/09 Applied a patch from the Sieve maintainer which:
3984 o fixes some comments
3985 o adds the (disabled) notify extension core
3986 o adds some debug output for the result of if/elsif tests
3987 o points to the current vacation draft in the documentation
3988 and documents the missing references header update
3992 o fixes a bug in processing the envelope test (when testing
3993 multiple envelope elements, the last element determined the
3996 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3997 Electronic Mail") by including:
3999 Auto-submitted: auto-generated
4001 in the messages that it generates (bounce messages and others, such as
4002 warnings). In the case of bounce messages for non-SMTP messages, there was
4003 also a typo: it was using "Auto_submitted" (underscore instead of
4004 hyphen). Since every message generated by Exim is necessarily in response
4005 to another message, thes have all been changed to:
4007 Auto-Submitted: auto-replied
4009 in accordance with these statements in the RFC:
4011 The auto-replied keyword:
4013 - SHOULD be used on messages sent in direct response to another
4014 message by an automatic process,
4016 - MUST NOT be used on manually-generated messages,
4018 - MAY be used on Delivery Status Notifications (DSNs) and Message
4019 Disposition Notifications (MDNs),
4021 - MUST NOT be used on messages generated by automatic or periodic
4022 processes, except for messages which are automatic responses to
4025 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4026 to the default Received: header definition.
4028 PH/12 Added log selector acl_warn_skipped (default on).
4030 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4031 variables because (a) they are in the wrong storage pool and (b) even if
4032 they were copied, it wouldn't work properly because of the caching.
4034 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4035 checking when decoding. Apparently there are clients that generate
4036 overlong encoded strings. Why am I not surprised?
4038 PH/15 If the first argument of "${if match_address" was not empty, but did not
4039 contain an "@" character, Exim crashed. Now it writes a panic log message
4040 and treats the condition as false.
4042 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4044 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4045 extension "envelope-auth". The code is finished and in agreement with
4046 other implementations, but there is no documentation so far and in fact,
4047 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4048 not changing the active code.
4050 Print executed "if" and "elsif" statements when debugging is used. This
4051 helps a great deal to understand what a filter does.
4053 Document more things not specified clearly in RFC3028. I had all this
4054 sorted out, when out of a sudden new issues came to my mind. Oops."
4056 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4059 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4060 canonical form (as documented). However, after a host name lookup from
4061 the IP address, check_host() was doing a simple string comparison with
4062 addresses acquired from the DNS when checking that the found name did
4063 have the original IP as one of its addresses. Since any found IPv6
4064 addresses are likely to be in abbreviated form, the comparison could
4065 fail. Luckily, there already exists a function for doing the comparison
4066 by converting both addresses to binary, so now that is used instead of
4067 the text comparison.
4069 PH/20 There was another similar case to PH/19, when a complete host name was
4070 given in a host list; looking up its IP address could give an abbreviated
4071 form, whereas the current host's name might or might not be abbreviated.
4072 The same fix has been applied.
4078 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4079 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4082 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4083 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4085 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4087 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4088 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4089 Also, for "Auto-Submitted", treat anything other than "no" as
4090 non-personal, in accordance with RFC 3834. (Previously it treated
4091 anything starting "auto-" as non-personal.)
4093 TF/01 The control=submission/name=... option had a problem with syntax
4094 errors if the name included a slash character. The /name= option
4095 now slurps the rest of the string, so it can include any characters
4096 but it must come last in the list of options (after /sender_retain
4099 PH/05 Some modifications to the interface to the fake nameserver for the new
4107 TK/01 Added the "success_on_redirect" address verification option. See
4108 NewStuff for rationale and an example.
4110 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4112 PH/02 Patch to exigrep to allow it to work on syslog lines.
4114 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4115 fread() to skip over the body file's header line, because in Cygwin the
4116 header line is locked and is inaccessible.
4118 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4119 co-exist for some time) to make it clear that it is the Exim ID that is
4120 referenced, not the Message-ID: header line.
4122 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4123 string_format() function, because snprintf() does not exist on all
4126 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4127 ${stat: expansion item.
4129 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4130 protocol synchronization error", to keep the pedants happy.
4132 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4133 well as for IRIX systems, when gcc is being used. See the host.c source
4136 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4138 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4141 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4142 transport and to the smtp transport in LMTP mode.
4144 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4146 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4147 run a filter in a subprocess. This could lead to confusion in subsequent
4148 lookups in the parent process. There should also be a search_tidyup() at
4149 the end of the subprocess.
4151 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4152 only if the host matched helo_try_verify_hosts, which caused the
4153 verification to occur when the EHLO/HELO command was issued. The ACL just
4154 tested the remembered result. Now, if a previous verification attempt has
4155 not happened, "verify = helo" does it there and then.
4157 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4159 TK/03 Fix log output including CR from clamd.
4161 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4162 exist provoked a memory error which could cause a segfault.
4164 PH/15 Installed PCRE 6.2
4166 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4168 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4169 of the problem. Specifically, suggested +O2 rather than +O1 for the
4172 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4174 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4175 record, the port was not correctly logged when the outgoing_port log
4176 selector was set (it logged the transort's default port).
4178 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4179 fallback_hosts, and "hosts" in the smtp transport.
4181 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4182 host errors such as "Connection refused".
4184 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4185 authentication with radiusclient 0.4.9:
4187 - Error returned from rc_read_config was caught wrongly
4188 - Username/password not passed on to radius server due to wrong length.
4190 The presumption is that some radiusclient API changes for 4.51/PH/17
4191 were not taken care of correctly. The code is still untested by me (my
4192 Linux distribution still has 0.3.2 of radiusclient), but it was
4193 contributed by a Radius user.
4195 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4196 expanding the "port" option of the smtp transport.
4198 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4199 while reading a MIME header. Thanks to Tom Hughes for a patch.
4201 PH/24 Include config.h inside local_scan.h so that configuration settings are
4204 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4205 that in an ACL for RCPT (for example), you can examine exactly what was
4208 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4209 commands, but it was not correctly comparing the address with the actual
4210 client host address. Thus, it would show the EHLO address in Received:
4211 header lines when this was not necessary.
4213 PH/27 Added the % operator to ${eval:}.
4215 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4216 it should be ignoring failures (because with -C, for example, it has lost
4217 privilege). It wasn't ignoring creation failures other than "already
4220 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4223 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4224 because an input line was too long, either on its own, or by virtue of
4225 too many continuations, the temporary file was not being removed, and the
4226 return code was incorrect.
4228 PH/31 Missing "BOOL" in function definition in filtertest.c.
4230 PH/32 Applied Sieve patches from the maintainer.
4232 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4234 PH/33 Added "verify = not_blind".
4236 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4237 Local/Makefile (with some defaults set). These are used in built scripts
4238 such as exicyclog, but they have never been used in the exim_install
4239 script (though there are many overriding facilities there). I have
4240 arranged that the exim_install script now takes note of these two
4243 PH/35 Installed configuration files for Dragonfly.
4245 PH/36 When a locally submitted message by a trusted user did not contain a
4246 From: header, and the sender address was obtained from -f or from an SMTP
4247 MAIL command, and the trusted user did not use -F to supply a sender
4248 name, $originator_name was incorrectly used when constructing a From:
4249 header. Furthermore, $originator_name was used for submission mode
4250 messages from external hosts without From: headers in a similar way,
4251 which is clearly wrong.
4253 PH/37 Added control=suppress_local_fixups.
4255 PH/38 When log_selector = +received_sender was set, and the addition of the
4256 sender made the log line's construction buffer exactly full, or one byte
4257 less than full, an overflow happened when the terminating "\n" was
4260 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4261 when the result of a list match is failure because a DNS lookup failed.
4263 PH/40 RM_COMMAND is now used in the building process.
4265 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4266 the "build-* directories that it finds.
4268 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4269 address in a domain literal was a prefix of an interface address.
4271 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4272 when verifying a sender address, unless rewrite_headers is false.
4274 PH/44 Wrote a long comment about why errors_to addresses are verified as
4275 recipients, not senders.
4277 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4278 the ratelimit ACL was added.
4280 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4282 PH/46 Added extra information about PostgreSQL errors to the error string.
4284 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4285 avoiding the need to install special zones in a real server. This is
4286 backwards compatible; if it can't find the fake resolver, it drops back.
4287 Thus, both old and new test suites can be run.
4289 TF/02 Added util/ratelimit.pl
4291 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4292 clock is set back in time.
4294 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4295 Candler <B.Candler@pobox.com>.
4297 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4298 OK for addresses that are the result of redirections.
4300 PH/48 A number of further additions for the benefit of the new test suite,
4301 including a fake gethostbyname() that interfaces to the fake DNS resolver
4304 TF/06 The fix for widen_domains has also been applied to qualify_single and
4305 search_parents which are the other dnslookup options that can cause
4308 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4311 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4313 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4314 service" if it encountered an MX record with an empty target hostname.
4315 The message is now "an MX or SRV record indicated no SMTP service".
4317 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4318 if the DNS of the sending site is misconfigured. This is quite a
4319 common situation. This change restores the behaviour of treating a
4320 helo verification defer as a failure.
4322 PH/51 If self=fail was set on a router, the bounce message did not include the
4323 actual error message.
4329 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4331 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4332 command itself ended in error, the underlying message about the transport
4333 filter timeout was being overwritten with the pipe command error. Now the
4334 underlying error message should be appended to the second error message.
4336 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4338 PH/02 Reduce the amount of output that "make" produces by default. Full output
4339 can still be requested.
4341 PH/03 The warning log line about a condition test deferring for a "warn" verb
4342 was being output only once per connection, rather than after each
4343 occurrence (because it was using the same function as for successful
4344 "warn" verbs). This seems wrong, so I have changed it.
4346 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4347 it should not have, which might have caused a crash in the right
4348 circumstances, but probably never did.
4350 PH/04 Installed a modified version of Tony Finch's patch to make submission
4351 mode fix the return path as well as the Sender: header line, and to
4352 add a /name= option so that you can make the user's friendly name appear
4355 TF/03 Added the control = fakedefer ACL modifier.
4357 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4358 Mark Lowes for thorough testing.
4360 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4362 TK/03 Merged latest SRS patch from Miles Wilton.
4364 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4365 with the definition in sysexits.h (which is #included earlier).
4366 Fortunately, Exim does not actually use EX_OK. The code used to try to
4367 preserve the sysexits.h value, by assuming that macro definitions were
4368 scanned for macro replacements. I have been disabused of this notion,
4369 so now the code just undefines EX_OK before #including unistd.h.
4371 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4372 in the smtp transport. When a block could not be written in a single
4373 write() function, the timeout was being re-applied to each part-write.
4374 This seems wrong - if the receiver was accepting one byte at a time it
4375 would take for ever. The timeout is now adjusted when this happens. It
4376 doesn't have to be particularly precise.
4378 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4379 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4381 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4382 without a domain if the check to <postmaster@domain> fails.
4384 SC/01 Eximstats: added -xls and the ability to specify output files
4385 (patch written by Frank Heydlauf).
4387 SC/02 Eximstats: use FileHandles for outputting results.
4389 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4391 SC/04 Eximstats: fixed display of large numbers with -nvr option
4393 SC/05 Eximstats: fixed merging of reports with empty tables.
4395 SC/06 Eximstats: added the -include_original_destination flag
4397 SC/07 Eximstats: removed tabs and trailing whitespace.
4399 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4401 TK/06 MBOX spool code: Add real "From " MBOX separator line
4402 so the .eml file is really in mbox format (even though
4403 most programs do not really care). Patch from Alex Miller.
4405 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4406 The latter is generated from $received_to and is only set if the
4407 message has one envelope recipient. SA can use these headers,
4408 obviously out-of-the-box. Patch from Alex Miller.
4410 PH/08 The ${def test on a variable was returning false if the variable's
4411 value was "0", contrary to what the specification has always said!
4412 The result should be true unless the variable is empty.
4414 PH/09 The syntax error of a character other than { following "${if
4415 def:variable_name" (after optional whitespace) was not being diagnosed.
4416 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4417 accidental colon was present, for example, could give incorrect results.
4419 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4420 result is used (not including appendfile, where other changes are about
4423 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4424 This involved changing a lot of size variables from int to off_t. It
4425 should work with maildirs and everything.
4427 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4428 spamd dying while we are connected to it.
4430 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4433 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4434 function for BDB 4.3.
4436 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4438 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4439 the sources for PCRE have been re-arranged and more files are now
4442 PH/15 The code I had for printing potentially long long variables in PH/11
4443 above was not the best (it lost precision). The length of off_t variables
4444 is now inspected at build time, and an appropriate printing format (%ld
4445 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4446 to be "long long int" or "long int". This is needed for the internal
4447 formatting function string_vformat().
4449 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4450 the configuration file to be ":syslog", then the script "guesses" where
4451 the logs files are, rather than using the compiled in default. In our
4452 case the guess is not the same as the compiled default, so the script
4453 suddenly stopped working when I started to use syslog. The patch checks
4454 to see if log_file_path is "". If so, it attempts to read it from exim
4455 with no configuration file to get the compiled in version, before it
4456 falls back to the previous guessing code."
4458 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4459 implementing BATV in an Exim configuration. See NewStuff for the gory
4462 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4463 Makefile that are specific to HP-UX.
4465 PH/18 If the "use_postmaster" option was set for a recipient callout together
4466 with the "random" option, the postmaster address was used as the MAIL
4467 FROM address for the random test, but not for the subsequent recipient
4468 test. It is now used for both.
4470 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4471 patch removes a few documentation additions to RFC 3028, because the
4472 latest draft now contains them. It adds the new en;ascii-case comparator
4473 and a new error check for 8bit text in MIME parts. Comparator and
4474 require names are now matched exactly. I enabled the subaddress
4475 extension, but it is not well tested yet (read: it works for me)."
4477 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4478 rework some of the code of TK/09 above to avoid the hardwired use of
4479 "%lld" and "long long". Replaced the call to snprintf() with a call to
4482 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4483 records point to non-existent hosts", "retry timeout exceeded", and
4484 "retry time not reached for any host after a long failure period".
4486 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4487 experimental DomainKeys support:
4489 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4490 (2) On an error such as an illegally used "control", the wrong name for
4491 the control was given.
4493 These problems did NOT occur unless DomainKeys support was compiled.
4495 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4497 PH/24 Added ${if match_ip condition.
4499 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4500 databases so that it will be absolutely obvious if a crash occurs in the
4501 DB library. This is a regular occurrence (often caused by mis-matched
4504 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4505 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4506 picked up on a user's system that detects such things. There doesn't seem
4507 to be a gcc warning option for this - only an attribute that has to be
4508 put on the function's prototype. It seems that in Fedora Core 4 they have
4509 set this on a number of new functions. No doubt there will be more in due
4512 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4513 specify a transport. However, if an address that was verified by such a
4514 router was the subject of a callout, Exim crashed because it tried to
4515 read the rcpt_include_affixes from the non-existent transport. Now it
4516 just assumes that the setting of that option is false. This bug was
4517 introduced by 4.51/PH/31.
4519 PH/28 Changed -d+all to exclude +memory, because that information is very
4520 rarely of interest, but it makes the output a lot bigger. People tend to
4521 do -d+all out of habit.
4523 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4524 code in os-type was giving problems when libc.so lives in lib64, like on
4527 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4528 aren't the modern standard, and it seems that some systems' include files
4529 don't always have them. Exim was already checking for some of the newer
4530 ones like T_AAAA, and defining it itself. I've added checks for all the
4531 record types that Exim uses.
4533 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4534 not automatically generating a new one, as it is supposed to. This
4535 prevented TLS from working. If the file did exist, but contained invalid
4536 data, a new version was generated, as expected. It was only the case of a
4537 non-existent file that was broken.
4539 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4540 with a change in libdomainkeys > 0.64.
4542 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4543 from DNS. If the selector record carries the flag, it now has
4544 precedence over the domain-wide flag.
4546 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4548 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4549 the use of a port name that isn't defined in /etc/services) occurred, the
4550 message was deferred as in a normal delivery, and thus remained on the
4551 spool, instead of being failed because of the mua_wrapper setting. This
4552 is now fixed, and I tidied up some of the mua_wrapper messages at the
4555 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4556 the messages in a hash of arrays rather than using individual hashes.
4557 This is a bit cleaner and results in dramatic memory savings, albeit
4558 at a slight CPU cost.
4560 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4561 as requested by Marc Sherman.
4563 SC/10 Eximstats: added histograms for user specified patterns as requested
4566 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4568 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4569 fopen() in the content-scanning modules that did not already have it.
4575 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4576 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4578 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4580 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4582 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4583 to test Sieve filters that use "vacation".
4585 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4586 that changes the way the GnuTLS parameters are stored in the cache file.
4587 The new format can be generated externally. For backward compatibility,
4588 if the data in the cache doesn't make sense, Exim assumes it has read an
4589 old-format file, and it generates new data and writes a new file. This
4590 means that you can't go back to an older release without removing the
4593 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4594 work if there are any delivery delays because "one_time" forces the
4595 parent to be marked "delivered", so its unseen clone is never tried
4596 again. For this reason, Exim now forbids the simultaneous setting of
4599 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4600 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4601 entry if you want to know the details. The fix, however, neglected to
4602 consider the case when local delivery batching is involved. The test for
4603 "previously delivered" was not happening when checking to see if an
4604 address could be batched with a previous (undelivered) one; under
4605 certain circumstances this could lead to multiple deliveries to the same
4608 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4609 in its include files, and this causes problems building Exim.
4611 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4612 header_syntax) but Exim was just ignoring anything given after a slash.
4613 In particular, this caused confusion with an attempt to use "verify =
4614 reverse_host_lookup/defer_ok". An error is now given when options are
4615 supplied for verify items that do not have them. (Maybe reverse_host_
4616 lookup should have a defer_ok option, but that's a different point.)
4618 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4619 defined by RFC 821) to 2048, because there were problems with some AUTH
4620 commands, and RFC 1869 says the size should be increased for extended
4621 SMTP commands that take arguments.
4623 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4626 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4627 "unknown" error; now it says that the functionality isn't in the binary.
4629 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4630 an address' error message when a string expansion fails (syntax or
4631 whatever). Otherwise the password may appear in the log. Following change
4632 PH/42 below, there is no longer a chance of it appearing in a bounce
4635 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4637 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4638 of its addresses was ever tried. (Bugzilla bug #2.)
4640 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4641 the result incorrectly in the debug output. (It correctly added a newline
4642 to what was transported.)
4644 TF/01 Added $received_time.
4646 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4647 commented out examples of how to interface to a virus scanner and to
4648 SpamAssassin. Also added commented examples of av_scanner and
4649 spamd_address settings.
4651 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4652 and controls are allowed in which ACLs. There were a couple of minor
4653 errors. Some of the entries in the conditions table (which is a table of
4654 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4655 negation of where the condition IS allowed.
4657 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4659 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4660 header file does not have a version number, so I've had to invent a new
4661 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4662 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4663 radiusclient), but it was contributed by a Radius user.
4665 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4666 files or directories, for OpenSSL.
4668 PH/19 When an Exim process that is running as root has to create an Exim log
4669 file, it does so in a subprocess that runs as exim:exim so as to get the
4670 ownership right at creation (otherwise, other Exim processes might see
4671 the file with the wrong ownership). There was no test for failure of this
4672 fork() call, which would lead to the process getting stuck as it waited
4673 for a non-existent subprocess. Forks do occasionally fail when resources
4674 run out. I reviewed all the other calls to fork(); they all seem to check
4677 PH/20 When checking for unexpected SMTP input at connect time (before writing
4678 the banner), Exim was not dealing correctly with a non-positive return
4679 from the read() function. If the client had disconnected by this time,
4680 the result was a log entry for a synchronization error with an empty
4681 string after "input=" when read() returned zero. If read() returned -1
4682 (an event I could not check), uninitialized data bytes were printed.
4683 There were reports of junk text (parts of files, etc) appearing after
4686 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4688 PH/22 Added support for macro redefinition, and (re)definition in between
4689 driver and ACL definitions.
4691 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4692 forgetting to use the resulting value; it was using the unexpanded value.
4694 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4695 hadn't been configured. The fix is from Juergen Kreileder, who
4696 understands it better than I do:
4698 "Here's what I see happening with three configured cyrus_sasl
4699 authenticators configured (plain, login, cram-md5):
4701 On startup auth_cyrus_sasl_init() gets called for each of these.
4702 This means three calls to sasl_listmech() without a specified mech_list.
4703 => SASL tests which mechs of all available mechs actually work
4704 => three warnings about OTP not working
4705 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4707 With the patch, sasl_listmech() also gets called three times. But now
4708 SASL's mech_list option is set to the server_mech specified in the the
4709 authenticator. Or in other words, the answer from sasl_listmech()
4710 gets limited to just the mech you're testing for (which is different
4712 => the return list contains just 'plain' or 'login', 'cram-md5' or
4713 nothing depending on the value of ob->server_mech.
4715 I've just tested the patch: Authentication still works fine,
4716 unavailable mechs specified in the exim configuration are still
4717 caught, and the auth.log warnings about OTP are gone."
4719 PH/25 When debugging is enabled, the contents of the command line are added
4720 to the debugging output, even when log_selector=+arguments is not
4723 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4724 answer is "GNU", and only if the return is "GNU/something" is the answer
4727 PH/27 $acl_verify_message is now set immediately after the failure of a
4728 verification in an ACL, and so is available in subsequent modifiers. In
4729 particular, the message can be preserved by coding like this:
4731 warn !verify = sender
4732 set acl_m0 = $acl_verify_message
4734 Previously, $acl_verify_message was set only while expanding "message"
4735 and "log_message" when a very denied access.
4737 PH/28 Modified OS/os.c-Linux with
4739 -#ifndef OS_LOAD_AVERAGE
4740 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4742 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4743 nomenclature these days.)
4745 PH/29 Installed patch from the Sieve maintainer that adds the options
4746 sieve_useraddress and sieve_subaddress to the redirect router.
4748 PH/30 In these circumstances:
4749 . Two addresses routed to the same list of hosts;
4750 . First host does not offer TLS;
4751 . First host accepts first address;
4752 . First host gives temporary error to second address;
4753 . Second host offers TLS and a TLS session is established;
4754 . Second host accepts second address.
4755 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4756 and peerdn, if requested) that were in fact used only for the second
4759 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4760 attention to any local part prefix or suffix that was matched by the
4761 router that accepted the address. It now behaves in the same way as it
4762 does for delivery: the affixes are removed from the local part unless
4763 rcpt_include_affixes is set on the transport.
4765 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4766 timeout during the DATA phase of an incoming message.
4768 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4769 applied a patch sanctioned by the Sieve maintainer.
4771 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4772 the uid or gid is negative. A case of a negative gid caused this to be
4773 noticed. The fix allows for either to be negative.
4775 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4776 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4779 PH/36 The change PH/12 above was broken. Fixed it.
4781 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4782 the grounds that routing the same address twice would always produce the
4783 same answer. This might have been true once, but it is certainly no
4784 longer true now. Routing a child address may depend on the previous
4785 routing that produced that child. Some complicated redirection strategies
4786 went wrong when messages had multiple recipients, and made Exim's
4787 behaviour dependent on the order in which the addresses were given.
4789 I have moved the duplicate checking until after the routing is complete.
4790 Exim scans the addresses that are assigned to local and remote
4791 transports, and removes any duplicates. This means that more work will be
4792 done, as duplicates will always all be routed, but duplicates are
4793 presumably rare, so I don't expect this is of any significance.
4795 For deliveries to pipes, files, and autoreplies, the duplicate checking
4796 still happens during the routing process, since they are not going to be
4799 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4800 It corrects a timeout issue with spamd. This is Ian's comment: "The
4801 background is that sometimes spamd either never reads data from a
4802 connection it has accepted, or it never writes response data. The exiscan
4803 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4804 blindly assumes that writes won't block so it may never time out."
4806 PH/39 Allow G after quota size as well as K and M.
4808 PH/40 The value set for $authenticated_id in an authenticator may not contain
4809 binary zeroes or newlines because the value is written to log lines and
4810 to spool files. There was no check on this. Now the value is run through
4811 the string_printing() function so that such characters are converted to
4812 printable escape sequences.
4814 PH/41 $message_linecount is a new variable that contains the total number of
4815 lines in the message. Compare $body_linecount, which is the count for the
4818 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4819 bounce and delay warning messages, except in certain special cases, which
4822 (a) An SMTP error message from a remote host;
4823 (b) A message specified in a :fail: redirection;
4824 (c) A message specified in a "fail" command in a system filter;
4825 (d) A message specified in a FAIL return from the queryprogram router;
4826 (e) A message specified by the cannot_route_message router option.
4828 In these cases only, Exim does include the error details in bounce and
4829 warning messages. There are also a few cases where bland messages such
4830 as "unrouteable address" or "local delivery error" are given.
4832 PH/43 $value is now also set for the "else" part of a ${run expansion.
4834 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4835 being worked on, but at least Exim now implements the latest version to
4838 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4839 process to complete was treated as a delivery failure, a timeout while
4840 writing the message to the pipe was logged, but erroneously treated as a
4841 successful delivery. Such timeouts include transport filter timeouts. For
4842 consistency with the overall process timeout, these timeouts are now
4843 treated as errors, giving rise to delivery failures by default. However,
4844 there is now a new Boolean option for the pipe transport called
4845 timeout_defer, which, if set TRUE, converts the failures into defers for
4846 both kinds of timeout. A transport filter timeout is now identified in
4849 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4850 systems where "make" and "gmake" are different, calling "gmake" at top
4851 level broke things. I've arranged for the value of $(MAKE) to be passed
4852 from the Makefile to this script so that it can call the same version of
4856 A note about Exim versions 4.44 and 4.50
4857 ----------------------------------------
4859 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4860 changes of various kinds. As a consequence, a big documentation update was
4861 needed. This delayed the release for rather longer than seemed good, especially
4862 in the light of a couple of (minor) security issues. Therefore, the changes
4863 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4864 release. So 4.44 and 4.50 are in effect two different branches that both start
4867 I have left the 4.50 change log unchanged; it contains all the changes since
4868 4.43. The change log for 4.44 is below; many of its items are identical to
4869 those for 4.50. This seems to be the most sensible way to preserve the
4870 historical information.
4876 1. Minor wording change to the doc/README.SIEVE file.
4878 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4879 computation of the current number of files was incorrect.
4881 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4882 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4885 4. Give more explanation in the error message when the command for a transport
4886 filter fails to execute.
4888 5. There are several places where Exim runs a non-Exim command in a
4889 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4890 was being done only for the command run by the queryprogram router. It is
4891 now done for all such subprocesses. The other cases are: ${run, transport
4892 filters, and the commands run by the lmtp and pipe transports.
4894 6. Added CONFIGURE_GROUP build-time option.
4896 7. Some older OS have a limit of 256 on the maximum number of file
4897 descriptors. Exim was using setrlimit() to set 1000 as a large value
4898 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4899 systems. I've change it so that if it can't get 1000, it tries for 256.
4901 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4902 was an oversight, and furthermore, ever since the addition of extra
4903 controls (e.g. 4.43/32), the checks on when to allow different forms of
4904 "control" were broken. There should now be diagnostics for all cases when a
4905 control that does not make sense is encountered.
4907 9. Added the /retain_sender option to "control=submission".
4909 10. $recipients is now available in the predata ACL (oversight).
4911 11. Tidy the search cache before the fork to do a delivery from a message
4912 received from the command line. Otherwise the child will trigger a lookup
4913 failure and thereby defer the delivery if it tries to use (for example) a
4914 cached ldap connection that the parent has called unbind on.
4916 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4917 of $address_data from the recipient verification was clobbered by the
4918 sender verification.
4920 13. The value of address_data from a sender verification is now available in
4921 $sender_address_data in subsequent conditions in the ACL statement.
4923 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4925 15. Added a new option "connect=<time>" to callout options, to set a different
4928 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4929 was its contents. (It was OK if the option was not defined at all.)
4931 17. A "Completed" log line is now written for messages that are removed from
4932 the spool by the -Mrm option.
4934 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4935 information about exactly what failed.
4937 19. Added -dd to debug only the daemon process.
4939 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4940 handles timeouts, both on the server side and network timeouts. Renamed the
4941 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4943 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4944 It is now set to "smtps".
4946 22. $host_address is now set to the target address during the checking of
4947 ignore_target_hosts.
4949 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4950 being passed; this would have caused $sender_host_name to have been used if
4951 matching the list had actually called for a host name (not very likely,
4952 since this list is usually IP addresses). A host name is now passed as
4955 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4956 code that specifies a non-restarting handler (typically sigaction() in
4957 modern systems) in an attempt to fix a rare and obscure crash bug.
4959 25. Narrowed the window for a race in the daemon that could cause it to ignore
4960 SIGCHLD signals. This is not a major problem, because they are used only to
4961 wake it up if nothing else does.
4963 26. A malformed maildirsize file could cause Exim to calculate negative values
4964 for the mailbox size or file count. Odd effects could occur as a result.
4965 The maildirsize information is now recalculated if the size or filecount
4968 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4969 support for a long time. Removed HAVE_SYS_VFS_H.
4971 28. Installed the latest version of exipick from John Jetmore.
4973 29. In an address list, if the pattern was not a regular expression, an empty
4974 subject address (from a bounce message) matched only if the pattern was an
4975 empty string. Non-empty patterns were not even tested. This was the wrong
4976 because it is perfectly reasonable to use an empty address as part of a
4977 database query. An empty address is now tested by patterns that are
4978 lookups. However, all the other forms of pattern expect the subject to
4979 contain a local part and a domain, and therefore, for them, an empty
4980 address still always fails if the pattern is not itself empty.
4982 30. Exim went into a mad DNS loop when attempting to do a callout where the
4983 host was specified on an smtp transport, and looking it up yielded more
4984 than one IP address.
4986 31. Re-factored the code for checking spool and log partition space into a
4987 function that finds that data and another that does the check. The former
4988 is then used to implement four new variables: $spool_space, $log_space,
4989 $spool_inodes, and $log_inodes.
4991 32. The RFC2047 encoding function was originally intended for short strings
4992 such as real names; it was not keeping to the 75-character limit for
4993 encoded words that the RFC imposes. It now respects the limit, and
4994 generates multiple encoded words if necessary. To be on the safe side, I
4995 have increased the buffer size for the ${rfc2047: expansion operator from
4998 33. It is now permitted to omit both strings after an "if" condition; if the
4999 condition is true, the result is "true". As before, when the second string
5000 is omitted, a false condition yields an empty string. This makes it less
5001 cumbersome to write custom ACL and router conditions.
5003 34. Failure to deliver a bounce message always caused it to be frozen, even if
5004 there was an errors_to setting on the router. The errors_to setting is now
5007 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5008 canonical form (fully expanded) before being placed in
5009 $sender_host_address.
5011 36. The table in the code that translates DNS record types into text (T_A to
5012 "A" for instance) was missing entries for NS and CNAME. It is just possible
5013 that this could have caused confusion if both these types were looked up
5014 for the same domain, because the text type is used as part of Exim's
5015 per-process caching. But the chance of anyone hitting this buglet seems
5018 37. The dnsdb lookup has been extended in a number of ways.
5020 (1) There is a new type, "zns", which walks up the domain tree until it
5021 finds some nameserver records. It should be used with care.
5023 (2) There is a new type, "mxh", which is like "mx" except that it returns
5024 just the host names, not the priorities.
5026 (3) It is now possible to give a list of domains (or IP addresses) to be
5027 looked up. The behaviour when one of the lookups defers can be
5028 controlled by a keyword.
5030 (4) It is now possible to specify the separator character for use when
5031 multiple records are returned.
5033 38. The dnslists ACL condition has been extended: it is now possible to supply
5034 a list of IP addresses and/or domains to be looked up in a particular DNS
5037 39. Added log_selector=+queue_time_overall.
5039 40. When running the queue in the test harness, wait just a tad after forking a
5040 delivery process, to get repeatability of debugging output.
5042 41. Include certificate and key file names in error message when GnuTLS fails
5043 to set them up, because the GnuTLS error message doesn't include the name
5044 of the failing file when there is a problem reading it.
5046 42. Allow both -bf and -bF in the same test run.
5048 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5050 44. The "Exiscan patch" is now merged into the mainline Exim source.
5052 45. Sometimes the final signoff response after QUIT could fail to get
5053 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5054 < 0 before doing a fflush(). This bug looks as though it goes back to the
5055 introduction of TLS in release 3.20, but "sometimes" must have been rare
5056 because the tests only now provoked it.
5058 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5059 (this can affect the format of dates).
5061 47. exim_tidydb, when checking for the continued existence of a message for
5062 which it has found a message-specific retry record, was not finding
5063 messages that were in split spool directories. Consequently, it was
5064 deleting retry records that should have stayed in existence.
5066 48. Steve fixed some bugs in eximstats.
5068 49. The SPA authentication driver was not abandoning authentication and moving
5069 on to the next authenticator when an expansion was forced to fail,
5070 contradicting the general specification for all authenticators. Instead it
5071 was generating a temporary error. It now behaves as specified.
5073 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5074 (the order specifies the preference for clients). The order is now AES256,
5075 AES128, 3DES, ARCFOUR128.
5077 51. Small patch to Sieve code - explicitly set From: when generating an
5080 52. Exim crashed if a remote delivery caused a very long error message to be
5081 recorded - for instance if somebody sent an entire SpamAssassin report back
5082 as a large number of 550 error lines. This bug was coincidentally fixed by
5083 increasing the size of one of Exim's internal buffers (big_buffer) that
5084 happened as part of the Exiscan merge. However, to be on the safe side, I
5085 have made the code more robust (and fixed the comments that describe what
5088 53. Now that there can be additional text after "Completed" in log lines (if
5089 the queue_time_overall log selector is set), a one-byte patch to exigrep
5090 was needed to allow it to recognize "Completed" as not the last thing in
5093 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5094 patch that reportedly fixes this has been added. I am not expert enough to
5095 create a test for it. This is what the patch creator wrote:
5097 "I found a little strange behaviour of ldap code when working with
5098 Windows 2003 AD Domain, where users was placed in more than one
5099 Organization Units. When I tried to give exim partial DN, the exit code
5100 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5101 But simultaneously result of request was absolutely normal ldap result,
5102 so I produce this patch..."
5104 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5105 REFERENCE, so I have modified the code to exclude the patch when that macro
5108 55. Some experimental protocols are using DNS PTR records for new purposes. The
5109 keys for these records are domain names, not reversed IP addresses. The
5110 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5111 leaves it alone. Component reversal etc. now happens only for IP addresses.
5114 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5116 57. Double the size of the debug message buffer (to 2048) so that more of very
5117 long debug lines gets shown.
5119 58. The exicyclog utility now does better if the number of log files to keep
5120 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5122 59. Two changes related to the smtp_active_hostname option:
5124 (1) $smtp_active_hostname is now available as a variable.
5125 (2) The default for smtp_banner uses $smtp_active_hostname instead
5126 of $primary_hostname.
5128 60. The host_aton() function is supposed to be passed a string that is known
5129 to be a valid IP address. However, in the case of IPv6 addresses, it was
5130 not checking this. This is a hostage to fortune. Exim now panics and dies
5131 if the condition is not met. A case was found where this could be provoked
5132 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5133 components; fortuitously, this particular loophole had already been fixed
5134 by change 4.50/55 above.
5136 If there are any other similar loopholes, the new check in host_aton()
5137 itself should stop them being exploited. The report I received stated that
5138 data on the command line could provoke the exploit when Exim was running as
5139 exim, but did not say which command line option was involved. All I could
5140 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5141 running as the user.
5144 61. There was a buffer overflow vulnerability in the SPA authentication code
5145 (which came originally from the Samba project). I have added a test to the
5146 spa_base64_to_bits() function which I hope fixes it.
5149 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5150 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5152 63. The daemon start-up calls getloadavg() while still root for those OS that
5153 need the first call to be done as root, but it missed one case: when
5154 deliver_queue_load_max is set with deliver_drop_privilege. This is
5155 necessary for the benefit of the queue runner, because there is no re-exec
5156 when deliver_drop_privilege is set.
5158 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5159 This has been fixed.
5161 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5162 was in use, was not putting the data itself into the right store pool;
5163 consequently, it could be overwritten for a subsequent message in the same
5164 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5167 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5169 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5170 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5171 the function were treating the return as a boolean value, which happened to
5172 work because 0=false and not-0=true, but is not correct code.
5174 68. The host_aton() function was not handling scoped IPv6 addresses (those
5175 with, for example, "%eth0" on the end) correctly.
5177 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5178 negated items (that is, ~something) in unsigned ints. Some compilers
5179 apparently mutter when there is no cast.
5181 70. If an address verification called from an ACL failed, and did not produce a
5182 user-specific message (i.e. there was only a "system" message), nothing was
5183 put in $acl_verify_message. In this situation, it now puts the system
5186 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5187 session; change 4.31/43 added the unwanted input to the log line - except
5188 that it did not do this in the start of session case. It now does.
5190 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5191 This is wrong and can cause the other end to generate a synchronization
5192 error if it is another Exim or anything else that does the synchronization
5193 check. A QUIT command is no longer sent after a timeout.
5195 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5196 during host lookups.
5198 74. The defer_ok option of callout verification was not working if it was used
5199 when verifying addresses in header lines, that is, for this case:
5201 verify = header_sender/callout=defer_ok
5203 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5204 those file descriptors could be used for SMTP connections. If anything
5205 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5206 could be sent to the SMTP client, causing chaos. The daemon now opens
5207 stdin, stdout, and stderr to /dev/null when it puts itself into the
5210 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5211 log by default. The user can override this with suitable Perl magic.
5213 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5214 the log message when discard triggers, was not working for the DATA ACL or
5215 for the non-SMTP ACL.
5217 78. Error message wording change in sieve.c.
5219 79. If smtp_accept_max_per_host was set, the number of connections could be
5220 restricted to fewer than expected, because the daemon was trying to set up
5221 a new connection before checking whether the processes handling previous
5222 connections had finished. The check for completed processes is now done
5223 earlier. On busy systems, this bug wouldn't be noticed because something
5224 else would have woken the daemon, and it would have reaped the completed
5227 80. If a message was submitted locally by a user whose login name contained one
5228 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5229 It caused a spool format error. I have fixed the spool reading code. A
5230 related problem was that the "from" clause in the Received: line became
5231 illegal because of the space(s). It is now covered by ${quote_local_part.
5233 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5236 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5238 83. Patch from Sieve maintainer to fix unterminated string problem in
5239 "vacation" handling.
5241 84. Some minor changes to the Linux configuration files to help with other
5242 OS variants using glibc.
5244 85. One more patch for Sieve to update vacation handling to latest spec.
5247 ----------------------------------------------------
5248 See the note above about the 4.44 and 4.50 releases.
5249 ----------------------------------------------------
5255 1. Change 4.43/35 introduced a bug that caused file counts to be
5256 incorrectly computed when quota_filecount was set in an appendfile
5259 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5260 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5263 3. Give more explanation in the error message when the command for a transport
5264 filter fails to execute.
5266 4. There are several places where Exim runs a non-Exim command in a
5267 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5268 was being done only for the command run by the queryprogram router. It is
5269 now done for all such subprocesses. The other cases are: ${run, transport
5270 filters, and the commands run by the lmtp and pipe transports.
5272 5. Some older OS have a limit of 256 on the maximum number of file
5273 descriptors. Exim was using setrlimit() to set 1000 as a large value
5274 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5275 systems. I've change it so that if it can't get 1000, it tries for 256.
5277 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5278 was an oversight, and furthermore, ever since the addition of extra
5279 controls (e.g. 4.43/32), the checks on when to allow different forms of
5280 "control" were broken. There should now be diagnostics for all cases when a
5281 control that does not make sense is encountered.
5283 7. $recipients is now available in the predata ACL (oversight).
5285 8. Tidy the search cache before the fork to do a delivery from a message
5286 received from the command line. Otherwise the child will trigger a lookup
5287 failure and thereby defer the delivery if it tries to use (for example) a
5288 cached ldap connection that the parent has called unbind on.
5290 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5291 of $address_data from the recipient verification was clobbered by the
5292 sender verification.
5294 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5295 was its contents. (It was OK if the option was not defined at all.)
5297 11. A "Completed" log line is now written for messages that are removed from
5298 the spool by the -Mrm option.
5300 12. $host_address is now set to the target address during the checking of
5301 ignore_target_hosts.
5303 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5304 being passed; this would have caused $sender_host_name to have been used if
5305 matching the list had actually called for a host name (not very likely,
5306 since this list is usually IP addresses). A host name is now passed as
5309 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5310 code that specifies a non-restarting handler (typically sigaction() in
5311 modern systems) in an attempt to fix a rare and obscure crash bug.
5313 15. Narrowed the window for a race in the daemon that could cause it to ignore
5314 SIGCHLD signals. This is not a major problem, because they are used only to
5315 wake it up if nothing else does.
5317 16. A malformed maildirsize file could cause Exim to calculate negative values
5318 for the mailbox size or file count. Odd effects could occur as a result.
5319 The maildirsize information is now recalculated if the size or filecount
5322 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5323 support for a long time. Removed HAVE_SYS_VFS_H.
5325 18. Updated exipick to current release from John Jetmore.
5327 19. Allow an empty sender to be matched against a lookup in an address list.
5328 Previously the only cases considered were a regular expression, or an
5331 20. Exim went into a mad DNS lookup loop when doing a callout where the
5332 host was specified on the transport, if the DNS lookup yielded more than
5335 21. The RFC2047 encoding function was originally intended for short strings
5336 such as real names; it was not keeping to the 75-character limit for
5337 encoded words that the RFC imposes. It now respects the limit, and
5338 generates multiple encoded words if necessary. To be on the safe side, I
5339 have increased the buffer size for the ${rfc2047: expansion operator from
5342 22. Failure to deliver a bounce message always caused it to be frozen, even if
5343 there was an errors_to setting on the router. The errors_to setting is now
5346 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5347 canonical form (fully expanded) before being placed in
5348 $sender_host_address.
5350 24. Updated eximstats to version 1.33
5352 25. Include certificate and key file names in error message when GnuTLS fails
5353 to set them up, because the GnuTLS error message doesn't include the name
5354 of the failing file when there is a problem reading it.
5356 26. Expand error message when OpenSSL has problems setting up cert/key files.
5359 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5360 (this can affect the format of dates).
5362 28. exim_tidydb, when checking for the continued existence of a message for
5363 which it has found a message-specific retry record, was not finding
5364 messages that were in split spool directories. Consequently, it was
5365 deleting retry records that should have stayed in existence.
5367 29. eximstats updated to version 1.35
5368 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5369 1.35 - bugfix such that pie charts by volume are generated correctly
5371 30. The SPA authentication driver was not abandoning authentication and moving
5372 on to the next authenticator when an expansion was forced to fail,
5373 contradicting the general specification for all authenticators. Instead it
5374 was generating a temporary error. It now behaves as specified.
5376 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5377 (the order specifies the preference for clients). The order is now AES256,
5378 AES128, 3DES, ARCFOUR128.
5380 31. Small patch to Sieve code - explicitly set From: when generating an
5383 32. Exim crashed if a remote delivery caused a very long error message to be
5384 recorded - for instance if somebody sent an entire SpamAssassin report back
5385 as a large number of 550 error lines. This bug was coincidentally fixed by
5386 increasing the size of one of Exim's internal buffers (big_buffer) that
5387 happened as part of the Exiscan merge. However, to be on the safe side, I
5388 have made the code more robust (and fixed the comments that describe what
5391 33. Some experimental protocols are using DNS PTR records for new purposes. The
5392 keys for these records are domain names, not reversed IP addresses. The
5393 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5394 leaves it alone. Component reversal etc. now happens only for IP addresses.
5397 34. The host_aton() function is supposed to be passed a string that is known
5398 to be a valid IP address. However, in the case of IPv6 addresses, it was
5399 not checking this. This is a hostage to fortune. Exim now panics and dies
5400 if the condition is not met. A case was found where this could be provoked
5401 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5402 components; fortuitously, this particular loophole had already been fixed
5403 by change 4.50/55 or 4.44/33 above.
5405 If there are any other similar loopholes, the new check in host_aton()
5406 itself should stop them being exploited. The report I received stated that
5407 data on the command line could provoke the exploit when Exim was running as
5408 exim, but did not say which command line option was involved. All I could
5409 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5410 running as the user.
5413 35. There was a buffer overflow vulnerability in the SPA authentication code
5414 (which came originally from the Samba project). I have added a test to the
5415 spa_base64_to_bits() function which I hope fixes it.
5418 36. The daemon start-up calls getloadavg() while still root for those OS that
5419 need the first call to be done as root, but it missed one case: when
5420 deliver_queue_load_max is set with deliver_drop_privilege. This is
5421 necessary for the benefit of the queue runner, because there is no re-exec
5422 when deliver_drop_privilege is set.
5424 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5425 was in use, was not putting the data itself into the right store pool;
5426 consequently, it could be overwritten for a subsequent message in the same
5427 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5430 38. Sometimes the final signoff response after QUIT could fail to get
5431 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5432 < 0 before doing a fflush(). This bug looks as though it goes back to the
5433 introduction of TLS in release 3.20, but "sometimes" must have been rare
5434 because the tests only now provoked it.
5440 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5441 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5442 Now it returns an integer. A number of calls were still expecting a T/F
5443 return. Fortuitously, in all cases, the tests worked in OK situations,
5444 which is the norm. However, things would have gone wrong on any write
5445 failures on the smtp file descriptor. This function is used when sending
5446 messages over SMTP and also when doing verify callouts.
5448 2. When Exim is called to do synchronous delivery of a locally submitted
5449 message (the -odf or -odi options), it no longer closes stderr before doing
5452 3. Implemented the mua_wrapper option.
5454 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5456 5. Implemented the functions header_remove(), header_testname(),
5457 header_add_at_position(), and receive_remove_recipient(), and exported them
5460 6. If an ACL "warn" statement specified the addition of headers, Exim already
5461 inserted X-ACL-Warn: at the start if there was no header name. However, it
5462 was not making this test for the second and subsequent header lines if
5463 there were newlines in the string. This meant that an invalid header could
5464 be inserted if Exim was badly configured.
5466 7. Allow an ACL "warn" statement to add header lines at the start or after all
5467 the Received: headers, as well as at the end.
5469 8. Added the rcpt_4xx retry error code.
5471 9. Added postmaster_mailfrom=xxx to callout verification option.
5473 10. Added mailfrom=xxxx to the callout verification option, for verify=
5476 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5477 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5479 12. Inserted some casts to stop certain compilers warning when using pointer
5480 differences as field lengths or precisions in printf-type calls (mostly
5481 affecting debugging statements).
5483 13. Added optional readline() support for -be (dynamically loaded).
5485 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5486 same clock tick as a message's arrival, so that its received time was the
5487 same as the "first fail" time on the retry record, and that message
5488 remained on the queue past the ultimate address timeout, every queue runner
5489 would try a delivery (because it was past the ultimate address timeout) but
5490 after another failure, the ultimate address timeout, which should have then
5491 bounced the address, did not kick in. This was a "< instead of <=" error;
5492 in most cases the first failure would have been in the next clock tick
5493 after the received time, and all would be well.
5495 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5496 being recognized when the domain list was tested by the match_domain
5497 condition in an expansion string.
5499 16. Added the ${str2b64: operator.
5501 17. Exim was always calling setrlimit() to set a large limit for the number of
5502 processes, without checking whether the existing limit was already
5503 adequate. (It did check for the limit on file descriptors.) Furthermore,
5504 errors from getrlimit() and setrlimit() were being ignored. Now they are
5505 logged to the main and panic logs, but Exim does carry on, to try to do its
5506 job under whatever limits there are.
5508 18. Imported PCRE 5.0.
5510 19. Trivial typo in log message " temporarily refused connection" (the leading
5513 20. If the log selector return_path_on_delivery was set and an address was
5514 redirected to /dev/null, the delivery process crashed because it assumed
5515 that a return path would always be set for a "successful" delivery. In this
5516 case, the whole delivery is bypassed as an optimization, and therefore no
5519 21. Internal re-arrangement: the function for sending a challenge and reading
5520 a response while authentication was assuming a zero-terminated challenge
5521 string. It's now changed to take a pointer and a length, to allow for
5522 binary data in such strings.
5524 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5526 23. Exim was not respecting finduser_retries when seeking the login of the
5527 uid under which it was called; it was always trying 10 times. (The default
5528 setting of finduser_retries is zero.) Also, it was sleeping after the final
5529 failure, which is pointless.
5531 24. Implemented tls_on_connect_ports.
5533 25. Implemented acl_smtp_predata.
5535 26. If the domain in control=submission is set empty, Exim assumes that the
5536 authenticated id is a complete email address when it generates From: or
5537 Sender: header lines.
5539 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5540 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5541 chown and chgrp in /bin and hostname in /usr/bin.
5543 28. Exim was keeping the "process log" file open after each use, just as it
5544 does for the main log. This opens the possibility of it remaining open for
5545 long periods when the USR1 signal hits a daemon. Occasional processlog
5546 errors were reported, that could have been caused by this. Anyway, it seems
5547 much more sensible not to leave this file open at all, so that is what now
5550 29. The long-running daemon process does not normally write to the log once it
5551 has entered its main loop, and it closes the log before doing so. This is
5552 so that log files can straightforwardly be renamed and moved. However,
5553 there are a couple of unusual error situations where the daemon does write
5554 log entries, and I had neglected to close the log afterwards.
5556 30. The text of an SMTP error response that was received during a remote
5557 delivery was being truncated at 512 bytes. This is too short for some of
5558 the long messages that one sometimes sees. I've increased the limit to
5561 31. It is now possible to make retry rules that apply only when a message has a
5562 specific sender, in particular, an empty sender.
5564 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5565 it possible to be selective about when SMTP synchronization is enforced.
5567 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5569 32. Implemented hosts_connection_nolog.
5571 33. Added an ACL for QUIT.
5573 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5576 35. Added mailbox_size and mailbox_filecount to appendfile.
5578 36. Added control = no_multiline_responses to ACLs.
5580 37. There was a bug in the logic of the code that waits for the clock to tick
5581 in the case where the clock went backwards by a substantial amount such
5582 that the microsecond fraction of "now" was more than the microsecond
5583 fraction of "then" (but the whole seconds number was less).
5585 38. Added support for the libradius Radius client library this is found on
5586 FreeBSD (previously only the radiusclient library was supported).
5592 1. When certain lookups returned multiple values in the form name=value, the
5593 quoting of the values was not always being done properly. Specifically:
5594 (a) If the value started with a double quote, but contained no whitespace,
5596 (b) If the value contained whitespace other than a space character (i.e.
5597 tabs or newlines or carriage returns) it was not quoted.
5598 This fix has been applied to the mysql and pgsql lookups by writing a
5599 separate quoting function and calling it from the lookup code. The fix
5600 should probably also be applied to nisplus, ibase and oracle lookups, but
5601 since I cannot test any of those, I have not disturbed their existing code.
5603 2. A hit in the callout cache for a specific address caused a log line with no
5604 reason for rejecting RCPT. Now it says "Previous (cached) callout
5605 verification failure".
5607 3. There was an off-by-one bug in the queryprogram router. An over-long
5608 return line was truncated at 256 instead of 255 characters, thereby
5609 overflowing its buffer with the terminating zero. As well as fixing this, I
5610 have increased the buffer size to 1024 (and made a note to document this).
5612 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5613 when Exim is waiting for an SMTP response from a remote server, Exim
5614 restarts its select() call on the socket, thereby resetting its timeout.
5615 This is not a problem when such interrupts are rare. Somebody set up a cron
5616 job to run exiwhat every 2 minutes, which is less than the normal select()
5617 timeout (5 or 10 minutes). This meant that the select() timeout never
5618 kicked in because it was always reset. I have fixed this by comparing the
5619 time when an interrupt arrives with the time at the start of the first call
5620 to select(). If more time than the timeout has elapsed, the interrupt is
5621 treated as a timeout.
5623 5. Some internal re-factoring in preparation for the addition of Sieve
5624 extensions (by MH). In particular, the "personal" test is moved to a
5625 separate function, and given an option for scanning Cc: and Bcc: (which is
5626 not set for Exim filters).
5628 6. When Exim created an email address using the login of the caller as the
5629 local part (e.g. when creating a From: or Sender: header line), it was not
5630 quoting the local part when it contained special characters such as @.
5632 7. Installed new OpenBSD configuration files.
5634 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5635 try to make them clearer.
5637 9. Callout options, other than the timeout value, were being ignored when
5638 verifying sender addresses in header lines. For example, when using
5640 verify = header_sender/callout=no_cache
5642 the cache was (incorrectly) being used.
5644 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5645 only the Cygwin environment.
5647 11. When return_path_on_delivery was set as a log selector, if different remote
5648 addresses in the same message used different return paths and parallel
5649 remote delivery occurred, the wrong values would sometimes be logged.
5650 (Whenever a remote delivery process finished, the return path value from
5651 the most recently started remote delivery process was logged.)
5653 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5654 lines when AUTH and/or TLS are in use. This is the "received protocol"
5655 field. Exim used to use "asmtp" for authenticated SMTP, without any
5656 indication (in the protocol name) for TLS use. Now it follows the RFC and
5657 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5658 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5659 names appear in log lines as well as in Received: header lines.
5661 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5662 extensions, and comparison tests, and to fix some bugs.
5664 14. Changes to the "personal" filter test:
5666 (1) The test was buggy in that it was just doing the equivalent of
5667 "contains" tests on header lines. For example, if a user's address was
5668 anne@some.where, the "personal" test would incorrectly be true for
5670 To: susanne@some.where
5672 This test is now done by extracting each address from the header in turn,
5673 and checking the entire address. Other tests that are part of "personal"
5674 are now done using regular expressions (for example, to check local parts
5675 of addresses in From: header lines).
5677 (2) The list of non-personal local parts in From: addresses has been
5678 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5679 taken from the Sieve specification recommendations.
5681 (3) If the message contains any header line starting with "List-" it is
5682 treated as non-personal.
5684 (4) The test for "circular" in the Subject: header line has been removed
5685 because it now seems ill-conceived.
5687 15. Minor typos in src/EDITME comments corrected.
5689 16. Installed latest exipick from John Jetmore.
5691 17. If headers_add on a router specified a text string that was too long for
5692 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5693 of string_sprintf() is now avoided.
5695 18. $message_body_size was not set (it was always zero) when running the DATA
5696 ACL and the local_scan() function.
5698 19. For the "mail" command in an Exim filter, no default was being set for
5699 the once_repeat time, causing a random time value to be used if "once" was
5700 specified. (If the value happened to be <= 0, no repeat happened.) The
5701 default is now 0s, meaning "never repeat". The "vacation" command was OK
5702 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5703 (I found it when inspecting the code).
5705 20. There is now an overall timeout for performing a callout verification. It
5706 defaults to 4 times the callout timeout, which applies to individual SMTP
5707 commands during the callout. The overall timeout applies when there is more
5708 than one host that can be tried. The timeout is checked before trying the
5709 next host. This prevents very long delays if there are a large number of
5710 hosts and all are timing out (e.g. when the network connections are timing
5711 out). The value of the overall timeout can be changed by specifying an
5712 additional sub-option for "callout", called "maxwait". For example:
5714 verify = sender/callout=5s,maxwait=20s
5716 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5717 to the end before writing, but this should make it even safer).
5719 22. Exim was forgetting that it had advertised PIPELINING for the second and
5720 subsequent messages on an SMTP connection. It was also not resetting its
5721 memory on STARTTLS and an internal HELO.
5723 23. When Exim logs an SMTP synchronization error within a session, it now
5724 records whether PIPELINING has been advertised or not.
5726 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5727 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5728 rather than long int.
5730 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5732 26. Added the never_mail option to autoreply.
5738 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5739 crash if the getsockname() call failed; this can happen if a connection is
5740 closed very soon after it is established. The problem was simply in the
5741 order in which certain operations were done, causing Exim to try to write
5742 to the SMTP stream before it had set up the file descriptor. The bug has
5743 been fixed by making things happen in the correct order.
5749 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5750 before the connection was closed, thus losing the rejection response.
5752 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5753 some early Solaris releases, but causes trouble in current releases where
5754 socklen_t is defined.
5756 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5759 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5762 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5763 that releasing the top of it at the end releases what was used for sub-
5764 expansions (unless the block got too big). However, discard this block if
5765 the first thing is a variable or header, so that we can use its block when
5766 it is dynamic (useful for very large $message_headers, for example).
5768 6. Lookups now cache *every* query, not just the most recent. A new, separate
5769 store pool is used for this. It can be recovered when all lookup caches are
5770 flushed. Lookups now release memory at the end of their result strings.
5771 This has involved some general refactoring of the lookup sources.
5773 7. Some code has been added to the store_xxx() functions to reduce the amount
5774 of flapping under certain conditions.
5776 8. log_incoming_interface used to affect only the <= reception log lines. Now
5777 it causes the local interface and port to be added to several more SMTP log
5778 lines, for example "SMTP connection from", and rejection lines.
5780 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5782 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5784 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5785 could be overwritten at the end of the current message (or the start of a
5786 new message if it was set in a HELO ACL). The value is now preserved for
5787 the duration of the SMTP connection.
5789 12. If a transport had a headers_rewrite setting, and a matching header line
5790 contained an unqualified address, that address was qualified, even if it
5791 did not match any rewriting rules. The underlying bug was that the values
5792 of the flags that permit the existence of unqualified sender and recipient
5793 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5794 non-local messages, and by -bnq for local messages) were not being
5795 preserved with the message after it was received.
5797 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5798 "next input=" as part of the text comprising the host identity instead of
5799 the correct text. The code was using the same buffer for two different
5800 strings. However, depending on which order the printing function evaluated
5801 its arguments, the bug did not always show up. Under Linux, for example, my
5802 test suite worked just fine.
5804 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5805 doesn't work with some older versions of Perl. It has been changed to "my",
5806 which in any case is probably the better facility to use.
5808 15. A really picky compiler found some instances of statements for creating
5809 error messages that either had too many or two few arguments for the format
5812 16. The size of the buffer for calls to the DNS resolver has been increased
5813 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5814 for addresses that have a lot of PTR records. This alleviates a problem; it
5815 does not fully solve it.
5817 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5818 buffer now truncates the list and logs the incident, which is the same
5819 action as happens when Exim is looking up a host name and its aliases.
5820 Previously in this situation something unpredictable would happen;
5821 sometimes it was "internal error: store_reset failed".
5823 18. If a server dropped the connection unexpectedly when an Exim client was
5824 using GnuTLS and trying to read a response, the client delivery process
5825 crashed while trying to generate an error log message.
5827 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5828 string, for example:
5830 warn message = H1: something\nH2: something
5832 the text was added as a single header line from Exim's point of view
5833 though it ended up OK in the delivered message. However, searching for the
5834 second and subsequent header lines using $h_h2: did not work. This has been
5835 fixed. Similarly, if a system filter added multiple headers in this way,
5836 the routers could not see them.
5838 20. Expanded the error message when iplsearch is called with an invalid key to
5839 suggest using net-iplsearch in a host list.
5841 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5842 ACLs are no longer actually imposed (and a message to that effect is
5845 22. If a "gecos" field in a passwd entry contained escaped characters, in
5846 particular, if it contained a \" sequence, Exim got it wrong when building
5847 a From: or a Sender: header from that name. A second bug also caused
5848 incorrect handling when an unquoted " was present following a character
5849 that needed quoting.
5851 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5852 was not being matched caselessly.
5854 24. Arranged for all hyphens in the exim.8 source to be escaped with
5857 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5858 properly. Recipient callout cache records were still being keyed to include
5859 the sender, even when use_sender was set false. This led to far more
5860 callouts that were necessary. The sender is no longer included in the key
5861 when use_sender is false.
5863 26. Added "control = submission" modifier to ACLs.
5865 27. Added the ${base62d: operator to decode base 62 numbers.
5867 28. dnsdb lookups can now access SRV records.
5869 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5870 the configuration file.
5872 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5873 (-v) mode. This makes the output for a verbose queue run more intelligible.
5875 31. Added a use_postmaster feature to recipient callouts.
5877 32. Added the $body_zerocount variable, containing the number of binary zero
5878 bytes in the message body.
5880 33. The time of last modification of the "new" subdirectory is now used as the
5881 "mailbox time last read" when there is a quota error for a maildir
5884 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5886 35. Added +ignore_unknown as a special item in host lists.
5888 36. Code for decoding IPv6 addresses in host lists is now included, even if
5889 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5890 address was recognized as an IP address, but was then not correctly decoded
5891 into binary, causing unexpected and incorrect effects when compared with
5898 1. Very minor rewording of debugging text in manualroute to say "list of
5899 hosts" instead of "hostlist".
5901 2. If verify=header_syntax was set, and a header line with an unqualified
5902 address (no domain) and a large number of spaces between the end of the
5903 name and the colon was received, the reception process suffered a buffer
5904 overflow, and (when I tested it) crashed. This was caused by some obsolete
5905 code that should have been removed. The fix is to remove it!
5907 3. When running in the test harness, delay a bit after writing a bounce
5908 message to get a bit more predictability in the log output.
5910 4. Added a call to search_tidyup() just before forking a reception process. In
5911 theory, someone could use a lookup in the expansion of smtp_accept_max_
5912 per_host which, without the tidyup, could leave open a database connection.
5914 5. Added the variables $recipient_data and $sender_data which get set from a
5915 lookup success in an ACL "recipients" or "senders" condition, or a router
5916 "senders" option, similar to $domain_data and $local_part_data.
5918 6. Moved the writing of debug_print from before to after the "senders" test
5921 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5922 problems for message scanning, either using a data ACL, or using
5923 local_scan() because the Received: header was not generated till after they
5924 were called (in order to set the time as the time of reception completion).
5925 I have revised the way this works. The header is now generated after the
5926 body is received, but before the ACL or local_scan() are called. After they
5927 are run, the timestamp in the header is updated.
5933 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5934 before starting a queue runner without re-exec. This happened only when
5935 deliver_drop_privilege was set or when the Exim user was set to root. The
5936 effect of the bug was that timeouts during subsequent deliveries caused
5937 crashes instead of being properly handled. The handler is now left at its
5938 default (and expected) setting.
5940 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5941 message, again when deliver_drop_privilege is set or Exim is run as root.
5942 The bug described in (1) was not present in this case, but the tidying up
5943 of the other signals was missing. I have made the two cases consistent.
5945 3. The ignore_target_hosts setting on a manualroute router was being ignored
5946 for hosts that were looked up using the /MX notation.
5948 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5951 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5952 operated on the sender address. After changing the $sender_address to <>
5953 for the sender address verify, Exim was re-instated it as the original
5954 (before rewriting) address, but remembering that it had rewritten it, so it
5955 wasn't rewriting it again. This bug also had the effect of breaking the
5956 sender address verification caching when the sender address was rewritten.
5958 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5959 This has been changed so that if the ip literal address matches
5960 ignore_target_hosts, the router declines.
5962 7. Added expansion conditions match_domain, match_address, and match_local_
5963 part (NOT match_host).
5965 8. The placeholder for the Received: header didn't have a length field set.
5967 9. Added code to Exim itself and to exim_lock to test for a specific race
5968 condition that could lead to file corruption when using MBX delivery. The
5969 issue is with the lockfile that is created in /tmp. If this file is removed
5970 after a process has opened it but before that process has acquired a lock,
5971 there is the potential for a second process to recreate the file and also
5972 acquire a lock. This could lead to two Exim processes writing to the file
5973 at the same time. The added code performs the same test as UW imapd; it
5974 checks after acquiring the lock that its file descriptor still refers to
5975 the same named file.
5977 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5978 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5981 11. Added the smtp_active_hostname option. If used, this will typically be made
5982 to depend on the incoming interface address. Because $interface_address is
5983 not set up until the daemon has forked a reception process, error responses
5984 that can happen earlier (such as "too many connections") no longer contain
5987 12. If an expansion in a condition on a "warn" statement fails because a lookup
5988 defers, the "warn" statement is abandoned, and the next ACL statement is
5989 processed. Previously this caused the whole ACL to be aborted.
5991 13. Added the iplsearch lookup type.
5993 14. Added ident_timeout as a log selector.
5995 15. Added tls_certificate_verified as a log selector.
5997 16. Added a global option tls_require_ciphers (compare the smtp transport
5998 option of the same name). This controls incoming TLS connections.
6000 17. I finally figured out how to make tls_require_ciphers do a similar thing
6001 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6002 before starting the TLS session.
6004 18. Tabs are now shown as \t in -bP output.
6006 19. If the log selector return_path_on_delivery was set, Exim crashed when
6007 bouncing a message because it had too many Received: header lines.
6009 20. If two routers both had headers_remove settings, and the first one included
6010 a superfluous trailing colon, the final name in the first list and the
6011 first name in the second list were incorrectly joined into one item (with a
6012 colon in the middle).
6018 1. Added -C and -D options to the exinext utility, mainly to make it easier
6019 to include in the automated testing, but these could be helpful when
6020 multiple configurations are in use.
6022 2. The exinext utility was not formatting the output nicely when there was
6023 an alternate port involved in the retry record key, nor when there was a
6024 message id as well (for retries that were specific to a specific message
6025 and a specific host). It was also confused by IPv6 addresses, because of
6026 the additional colons they contain. I have fixed the IPv4 problem, and
6027 patched it up to do a reasonable job for IPv6.
6029 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6030 delivery, the log line now contains "pipelined" if PIPELINING was used.
6032 4. An SMTP transport process used to panic and die if the bind() call to set
6033 an explicit outgoing interface failed. This has been changed; it is now
6034 treated in the same way as a connect() failure.
6036 5. A reference to $sender_host_name in the part of a conditional expansion
6037 that was being skipped was still causing a DNS lookup. This no longer
6040 6. The def: expansion condition was not recognizing references to header lines
6041 that used bh_ and bheader_.
6043 7. Added the _cache feature to named lists.
6045 8. The code for checking quota_filecount in the appendfile transport was
6046 allowing one more file than it should have been.
6048 9. For compatibility with Sendmail, the command line option
6056 and sets the incoming protocol and host name (for trusted callers). The
6057 host name and its colon can be omitted when only the protocol is to be set.
6058 Note the Exim already has two private options, -pd and -ps, that refer to
6059 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6060 "s", but I don't think that's a major issue.
6062 10. A number of refactoring changes to the code, none of which should affect
6065 (a) The number of logging options was getting close to filling up the
6066 32-bit word that was used as a bit map. I have split them into two classes:
6067 those that are passed in the argument to log_write(), and those that are
6068 only ever tested independently outside of that function. These are now in
6069 separate 32-bit words, so there is plenty of room for expansion again.
6070 There is no change in the user interface or the logging behaviour.
6072 (b) When building, for example, log lines, the code previously used a
6073 macro that called string_cat() twice, in order to add two strings. This is
6074 not really sufficiently general. Furthermore, there was one instance where
6075 it was actually wrong because one of the argument was used twice, and in
6076 one call a function was used. (As it happened, calling the function twice
6077 did not affect the overall behaviour.) The macro has been replaced by a
6078 function that can join an arbitrary number of extra strings onto a growing
6081 (c) The code for expansion conditions now uses a table and a binary chop
6082 instead of a serial search (which was left over from when there were very
6083 few conditions). Also, it now recognizes conditions like "pam" even when
6084 the relevant support is not compiled in: a suitably worded error message is
6085 given if an attempt is made to use such a condition.
6087 11. Added ${time_interval:xxxxx}.
6089 12. A bug was causing one of the ddress fields not to be passed back correctly
6090 from remote delivery subprocesses. The field in question was not being
6091 subsequently used, so this caused to problems in practice.
6093 13. Added new log selectors queue_time and deliver_time.
6095 14. Might have fixed a bug in maildirsizefile handling that threw up
6096 "unexpected character" debug warnings, and recalculated the data
6097 unnecessarily. In any case, I expanded the warning message to give more
6100 15. Added the message "Restricted characters in address" to the statements in
6101 the default ACL that block characters like @ and % in local parts.
6103 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6104 Three changes have been made:
6106 (a) There was a serious bug; a negative response to MAIL caused the whole
6107 recipient domain to be cached as invalid, thereby blocking all messages
6108 to all local parts at the same domain, from all senders. This bug has
6109 been fixed. The domain is no longer cached after a negative response to
6110 MAIL if the sender used is not empty.
6112 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6115 (c) A new callout option, "use_sender" has been added for people who want
6116 the modified behaviour.
6122 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6125 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6126 indeed breaks things for older releases.
6128 3. Added additional logging to the case where there is a problem reading data
6129 from a filter that is running in a subprocess using a pipe, in order to
6130 try to track down a specific problem.
6132 4. Testing facility fudge: when running in the test harness and attempting
6133 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6134 getting "No route to host". Convert this to a timeout.
6136 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6139 6. Some OS don't have socklen_t but use size_t instead. This affects the
6140 fifth argument of getsockopt() amongst other things. This is now
6141 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6142 can be set for individual OS. I have set it for SunOS5, OSF1, and
6143 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6144 some earlier ones do not.
6146 7. Change 4.30/15 was not doing the test caselessly.
6148 8. The standard form for an IPv6 address literal was being rejected by address
6149 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6150 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6151 this, as well as the form without the "IPv6" on the front (but only when
6152 address literals are enabled, of course).
6154 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6156 10. Exim crashed if a message with an empty sender address specified by -f
6157 encountered a router with an errors_to setting. This could be provoked only
6158 by a command such as
6162 where an empty string was supplied; "<>" did not hit this bug.
6164 11. Installed PCRE release 4.5.
6166 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6167 remained set. It is now erased.
6169 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6170 times from message ids (which are base 36 rather than the normal 62).
6172 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6173 were being counted as actual protocol errors, and logged if the log
6174 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6175 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6176 and DATA following a set of rejected RCPTs do not count as protocol errors.
6177 In other words, Exim assumes they were pipelined, though this may not
6178 actually be the case. Of course, in all cases the client gets an
6179 appropriate error code.
6181 15. If a lookup fails in an ACL condition, a message about the failure may
6182 be available; it is used if testing the ACL cannot continue, because most
6183 such messages specify what the cause of the deferral is. However, some
6184 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6185 that caused an old message to be retained and used if a later statement
6186 caused a defer, replacing the real cause of the deferral.
6188 16. If an IP address had so many PTR records that the DNS lookup buffer
6189 was not large enough to hold them, Exim could crash while trying to process
6190 the truncated data. It now detects and logs this case.
6192 17. Further to 4.21/58, another change has been made: if (and only if) the
6193 first line of a message (the first header line) ends with CRLF, a bare LF
6194 in a subsequent header line has a space inserted after it, so as not to
6195 terminate the header.
6197 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6198 unnecessarily, used atoi() instead of strtol(), and didn't check the
6199 termination when getting file sizes from file names by regex.
6201 19. Completely re-implemented the support for maildirsize files, in the light
6202 of a number of problems with the previous contributed implementation
6203 (4.30/29). In particular:
6205 . If the quota is zero, the maildirsize file is maintained, but no quota is
6208 . If the maildir directory does not exist, it is created before any attempt
6209 to write a maildirsize file.
6211 . The quota value in the file is just a cache; if the quota is changed in
6212 the transport, the new value overrides.
6214 . A regular expression is available for excluding directories from the
6217 20. The autoreply transport checks the characters in options that define the
6218 message's headers; it allows continued headers, but it was checking with
6219 isspace() after an embedded newline instead of explicitly looking for a
6222 21. If all the "regular" hosts to which an address was routed had passed their
6223 expiry times, and had not reached their retry times, the address was
6224 bounced, even if fallback hosts were defined. Now Exim should go on to try
6227 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6228 equivalent code in the SMTP transport. Some hosts send humungous responses
6229 to HELO/EHLO, more than 1024 it seems.
6231 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6232 gives compiler warnings in some environments. I've now done it "properly",
6235 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6236 (because of problems with the built-in one) was declared to return uschar *
6237 instead of char *, causing compiler failure.
6239 25. Fixed a file descriptor leak when processing alias/forward files.
6241 26. Fixed a minor format string issue in dbfn.c.
6243 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6245 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6246 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6249 29. When a sender address is verified, it is cached, to save repeating the test
6250 when there is more than one recipient in a message. However, when the
6251 verification involves a callout, it is possible for different callout
6252 options to be set for different recipients. It is too complicated to keep
6253 track of this in the cache, so now Exim always runs a verification when a
6254 callout is required, relying on the callout cache for the optimization.
6255 The overhead is duplication of the address routing, but this should not be
6258 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6259 to be verified with callout=postmaster, and the main callout worked but the
6260 postmaster check failed, the verification correctly failed. However, if a
6261 subsequent RCPT command asked for sender verification *without* the
6262 postmaster check, incorrect caching caused this verification also to fail,
6265 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6266 it was not caching the DNS options (qualify_single, search_parents) that
6267 were used when the lookup failed. A subsequent lookup with different
6268 options therefore always gave the same answer, though there were cases
6269 where it should not have. (Example: a "domains = !$mx_any" option on a
6270 dnslookup router: the "domains" option is always processed without any
6271 widening, but the router might have qualify_single set.) Now Exim uses the
6272 cached value only when the same options are set.
6274 32. Added John Jetmore's "exipick" utility to the distribution.
6276 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6277 than a timeout (e.g. a certificate is required, and is not provided), an
6278 Exim server now closes the connection immediately. Previously it waited for
6279 the client to close - but if the client is SSL, it seems that they each
6280 wait for each other, leading to a delay before one of them times out.
6282 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6283 maintained 0.8.x compatibility because I don't think many are using it, and
6284 it is clearly obsolete.
6286 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6289 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6290 client certificate was expired. A simple patch fixes this, though I don't
6291 understand the full logic of why the verify callback is called multiple
6294 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6295 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6296 which causes problems with some clients (such as the Certicom SSL Plus
6297 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6298 disables the coutermeasure allowing Eudora to connect."
6300 38. Exim was not checking that a write() to a log file succeeded. This could
6301 lead to Bad Things if a log got too big, in particular if it hit a file
6302 size limit. Exim now panics and dies if it cannot write to a log file, just
6303 as it does if it cannot open a log file.
6305 39. Modified OS/Makefile-Linux so that it now contains
6307 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6309 The two -D definitions ensure that Exim is compiled with large file
6310 support, which makes it possible to handle log files that are bigger than
6313 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6314 instance) a domain was checked against a named list that involved a lookup,
6315 causing $domain_data to be set, then another domain was checked against the
6316 same list, then the first domain was re-checked, the value of $domain_data
6317 after the final check could be wrong. In particular, if the second check
6318 failed, it could be set empty. This bug probably also applied to
6321 41. The strip_trailing_dot option was not being applied to the address given
6322 with the -f command-line option.
6324 42. The code for reading a message's header from the spool was incrementing
6325 $received_count, but never initializing it. This meant that the value was
6326 incorrect (doubled) while delivering a message in the same process in which
6327 it was received. In the most common configuration of Exim, this never
6328 happens - a fresh exec is done - but it can happen when
6329 deliver_drop_privilege is set.
6331 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6332 it now includes up to 150 characters of the unexpected data in the log
6335 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6336 and building data strings. The size of both of these buffers was 10 000
6337 bytes - far larger than anybody would *ever* want, thought I. Needless to
6338 say, somebody hit the limit. I have increased the maximum line length to
6339 20 000 and the maximum data length of concatenated lines to 100 000. I have
6340 also fixed two bugs, because there was no checking on these buffers. Tsk,
6341 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6342 buffer is too small.
6344 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6345 lsearch lookups. Now it does.
6347 46. When parsing a route_list item in a manualroute router, a fixed-length
6348 buffer was used for the list of hosts. I made this 1024 bytes long,
6349 thinking that nobody would ever have a list of hosts that long. Wrong.
6350 Somebody had a whole pile of complicated expansion conditions, and the
6351 string was silently truncated, leading to an expansion error. It turns out
6352 that it is easier to change to an unlimited length (owing to other changes
6353 that have happened since this code was originally written) than to build
6354 structure for giving a limitation error. The length of the item that
6355 expands into the list of hosts is now unlimited.
6357 47. The lsearch lookup could not handle data where the length of text line was
6358 more than 4095 characters. Such lines were truncated, leading to shortened
6359 data being returned. It should now handle lines of any length.
6361 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6362 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6365 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6366 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6367 no longer happens in comment lines. A list of replacements is now placed
6368 at the head of all of the source files, except those whose only change is
6369 to replace PERL_COMMAND in the very first #! line.
6371 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6372 messages on the queue, with a bottom-up merge sort, using code contributed
6373 by Michael Haardt. This should make operations like -bp somewhat faster on
6374 large queues. It won't affect queue runners, except when queue_run_in_order
6377 51. Installed eximstats 1.31 in the distribution.
6379 52. Added support for SRV lookups to the dnslookup router.
6381 53. If an ACL referred to $message_body or $message_body_end, the value was not
6382 reset for any messages that followed in the same SMTP session.
6384 54. The store-handling optimization for building very long strings was not
6385 differentiating between the different store pools. I don't think this
6386 actually made any difference in practice, but I've tidied it.
6388 55. While running the routers to verify a sender address, $sender_address
6389 was still set to the sender address. This is wrong, because when routing to
6390 send a bounce to the sender, it would be empty. Therefore, I have changed
6391 it so that, while verifying a sender address, $sender_address is set to <>.
6392 (There is no change to what happens when verifying a recipient address.)
6394 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6395 target A or AAAA records (if not already returned) without resetting the
6396 qualify_single or search_parents options of the DNS resolver. These are
6397 inappropriate in this case because the targets of MX and SRV records must
6398 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6399 target that, when qualified, matched something in the local domain. These
6400 two options are now turned off when doing these lookups.
6402 57. It seems that at least some releases of Reiserfs (which does not have the
6403 concept of a fixed number of inodes) returns zero and not -1 for the
6404 number of available inodes. This interacted badly with check_spool_inodes,
6405 which assumed that -1 was the "no such thing" setting. What I have done is
6406 to check that the total number of inodes is greater than zero before doing
6407 the test of how many are available.
6409 58. When a "warn" ACL statement has a log_message modifier, the message is
6410 remembered, and not repeated. This is to avoid a lot of repetition when a
6411 message has many recipients that cause the same warning to be written.
6412 However, Exim was preserving the list of already written lines for an
6413 entire SMTP session, which doesn't seem right. The memory is now reset if a
6414 new message is started.
6416 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6417 debugging output unless log_rewrite was also set.
6419 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6420 of (int)(handle) when we know that handle contains (void *)(-1).
6422 61. The Exim daemon panic-logs an error return when it closes the incoming
6423 connection. However "connection reset by peer" seems to be common, and
6424 isn't really an error worthy of noting specially, so that particular error
6427 62. When Exim is trying to find all the local interfaces, it used to panic and
6428 die if the ioctl to get the interface flags failed. However, it seems that
6429 on at least one OS (Solaris 9) it is possible to have an interface that is
6430 included in the list of interfaces, but for which you get a failure error
6431 for this call. This happens when the interface is not "plumbed" into a
6432 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6433 failure of the "get flags" call assumes that the interface is down.
6435 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6436 makes life easier for people who are doing arithmetic on fields extracted
6437 from dates, where you often get leading zeros that should not be
6438 interpreted as octal.
6440 64. Added qualify_domain to the redirect router, to override the global
6443 65. If a pathologically long header line contained very many addresses (the
6444 report of this problem mentioned 10 000) and each of them was rewritten,
6445 Exim could use up a very large amount of memory. (It kept on making new
6446 copies of the header line as it rewrote, and never released the old ones.)
6447 At the expense of a bit more processing, the header rewriting function has
6448 been changed so that it no longer eats memory in this way.
6450 66. The generation of the Received: header has been moved from the time that a
6451 message starts to be received, to the time that it finishes. The timestamp
6452 in the Received: header should now be very close to that of the <= log
6453 line. There are two side-effects of this change:
6455 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6456 logged header lines no longer include the local Received: line, because
6457 it has not yet been created. The same applies to a copy of the message
6458 that is returned to a non-SMTP sender when a message is rejected.
6460 (b) When a filter file is tested using -bf, no additional Received: header
6461 is added to the test message. After some thought, I decided that this
6464 This change does not affect the value of $received_for. It is still set
6465 after address rewriting, but before local_scan() is called.
6467 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6469 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6470 gave an unhelpful panic error message, and a defer error. I have managed to
6471 change this behaviour so that it now rejects any supplied certificate,
6472 which seems right, as the list of acceptable certificates is empty.
6474 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6475 gave an unhelpful defer error. I have not managed to make this reject any
6476 supplied certificates, but the error message it gives is "no certificate
6477 supplied", which is not helpful.
6479 70. exigrep's output now also includes lines that are not associated with any
6480 message, but which match the given pattern. Implemented by a patch from
6481 Martin Sluka, which also tidied up the Perl a bit.
6483 71. Recipient callout verification, like sender verification, was using <> in
6484 the MAIL FROM command. This isn't really the right thing, since the actual
6485 sender may affect whether the remote host accepts the recipient or not. I
6486 have changed it to use the actual sender in the callout; this means that
6487 the cache record is now keyed on a recipient/sender pair, not just the
6488 recipient address. There doesn't seem to be a real danger of callout loops,
6489 since a callout by the remote host to check the sender would use <>.
6490 [SEE ABOVE: changed after hitting problems.]
6492 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6493 temporary errors. However, in the case of such a code being given after
6494 the end of a data transmission (i.e. after ".") Exim was failing to write
6495 a retry record for the message. (Yes, there was some broken host that was
6496 actually sending 8xx at this point.)
6498 73. An unknown lookup type in a host list could cause Exim to panic-die when
6499 the list was checked. (An example that provoked this was putting <; in the
6500 middle of a list instead of at the start.) If this happened during a DATA
6501 ACL check, a -D file could be left lying around. This kind of configuration
6502 error no longer causes Exim to die; instead it causes a defer error. The
6503 incident is still logged to the main and panic logs.
6505 74. Buglet left over from Exim 3 conversion. The message "too many messages
6506 in one connection" was written to the rejectlog but not the mainlog, except
6507 when address rewriting (yes!) was being logged.
6509 75. Added write_rejectlog option.
6511 76. When a system filter was run not as root (that is, when system_filter_user
6512 was set), the values of the $n variables were not being returned to the
6513 main process; thus, they were not subsequently available in the $sn
6516 77. Added +return_path_on_delivery log selector.
6518 78. A connection timeout was being treated differently from recipients deferred
6519 when testing hosts_max_try with a message that was older than the host's
6520 retry timeout. (The host should not be counted, thus allowing all hosts to
6521 be tried at least once before bouncing.) This may have been the cause of an
6522 occasionally reported bug whereby a message would remain on the queue
6523 longer than the retry timeout, but would be bounced if a delivery was
6524 forced. I say "may" because I never totally pinned down the problem;
6525 setting up timeout/retry tests is difficult. See also the next item.
6527 79. The ultimate address timeout was not being applied to errors that involved
6528 a combination of host plus message (for example, a timeout on a MAIL
6529 command). When an address resolved to a number of possible hosts, and they
6530 were not all tried for each delivery (e.g. because of hosts_max_try), a
6531 message could remain on the queue longer than the retry timeout.
6533 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6536 81. Fixed an obscure SMTP outgoing bug which required at least the following
6537 conditions: (a) there was another message waiting for the same server;
6538 (b) the server returned 5xx to all RCPT commands in the first message so
6539 that the message was not completed; (c) the server dropped the connection
6540 or gave a negative response to the RSET that Exim sends to abort the
6541 transaction. The observed case was a dropped connection after DATA that had
6542 been sent in pipelining mode. That is, the server had advertised PIPELINING
6543 but was not implementing it correctly. The effect of the bug was incorrect
6544 behaviour, such as trying another host, and this could lead to a crash.
6550 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6551 and daemon.c were passed as pointers to ints; they should have been
6552 pointers to socklen_t variables (which are typically unsigned ints).
6554 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6557 3. Fixed a really odd bug that affected only the testing scheme; patching a
6558 certain fixed string in the binary changed the value of another string that
6559 happened to be identical to the end of the original first string.
6561 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6562 name", it returns that address as the IP address. On some operating
6563 systems (e.g. Solaris), it also passes back the IP address string as the
6564 "host name". However, on others (e.g. Linux), it passes back an empty
6565 string. Exim wasn't checking for this, and was changing the host name to an
6566 empty string, assuming it had been canonicalized.
6568 5. Although rare, it is permitted to have more than one PTR record for a given
6569 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6570 all the names associated with an address, because they do in Solaris.
6571 However, it seems that they do not in Linux for data that comes from the
6572 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6573 I found this out when I moved to a new Linux workstation and tried to run
6574 the Exim test suite.
6576 To get round this problem I have changed the code so that it now does its
6577 own call to the DNS to look up PTR records when searching for a host name.
6578 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6579 addresses that are only in /etc/hosts are still found.
6581 This behaviour is, however, controlled by an option called host_lookup_
6582 order, which defaults to "bydns:byaddr". If people want to use the other
6583 order, or indeed, just use one or the other means of lookup, they can
6584 specify it in this variable.
6586 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6587 some operating systems, this comes back from gethostbyaddr() as an empty
6588 string, and this is what Exim used to test for. However, it seems that in
6589 other systems, "." is yielded. Exim now tests for this case too.
6591 7. The values of check_spool_space and check_log_space are now held internally
6592 as a number of kilobytes instead of an absolute number of bytes. If a
6593 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6594 kilobyte. This means that much larger values can be stored.
6596 8. Exim monitor: an attempt to get the action menu when not actually pointing
6597 at a message produces an empty menu entitled "No message selected". This
6598 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6599 no entries in it ("Shell widget menu has zero width and/or height"). So I
6600 have added a single, blank menu entry in this case.
6602 9. Added ${quote_local_part.
6604 10. MIME decoding is now applied to the contents of Subject: header lines when
6607 11. Now that a reference to $sender_host_address automatically causes a reverse
6608 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6609 host lookup before query-style lookups in lists that might use this
6610 variable. This has therefore been abolished, and the "net-" prefix is no
6611 longer necessary for query-style lookups.
6613 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6614 have been a typo for LFLAGS, so it has been changed.
6616 13. The install script calls Exim with "-C /dev/null" in order to find the
6617 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6618 to be output. However, since Exim outputs its version number before the
6619 error, it didn't break the script. It just looked ugly. I fixed this by
6620 always allowing "-C /dev/null" if the caller is root.
6622 14. Ignore overlarge ACL variable number when reading spool file - insurance
6623 against a later release with more variables having written the file.
6625 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6626 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6627 this, as well as the form without the "IPv6" on the front.
6629 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6630 OS/Makefile-Darwin file.
6632 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6633 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6635 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6636 variable. However, some OS use ints for inodes. I've added cast to long int
6637 to get rid of the compiler warning.
6639 19. I had forgotten to lock out "/../" in configuration file names when
6640 ALT_CONFIG_PREFIX was set.
6642 20. Routers used for verification do not need to specify transports. However,
6643 if such a router generated a host list, and callout was configured, Exim
6644 crashed, because it could not find a port number from the (non-existent)
6645 transport. It now assumes port 25 in this circumstance.
6647 21. Added the -t option to exigrep.
6649 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6650 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6651 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6652 them is set, LOOKUP_LSEARCH is forced.
6654 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6656 24. Added sender and host information to the "rejected by local_scan()" log
6657 line; previously there was no indication of these.
6659 25. Added .include_if_exists.
6661 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6662 Linux. It turns out that not all file systems support this. Apparently some
6663 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6664 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6665 not supported on the file descriptor, is now ignored when Exim is trying to
6666 sync a directory. This applies only to Linux.
6668 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6670 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6671 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6674 29. Added support for maildirsize files from supplied patch (modified a bit).
6676 30. The use of :fail: followed by an empty string could lead Exim to respond to
6677 sender verification failures with (e.g.):
6679 550 Verification failed for <xxx>
6680 550 Sender verify failed
6682 where the first response line was missing the '-' that indicates it is not
6683 the final line of the response.
6685 31. The loop for finding the name of the user that called Exim had a hardwired
6686 limit of 10; it now uses the value of finduser_retries, which is used for
6687 all other user lookups.
6689 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6692 33. Exim was neglecting to zero errno before one call of strtol() when
6693 expanding a string and expecting an integer value. On some systems this
6694 resulted in spurious "integer overflow" errors. Also, it was casting the
6695 result into an int without checking.
6697 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6698 did not work. The code looks as if it has *never* worked, though it appears
6699 to have been documented since at least release 1.62. I have made it work.
6701 35. The "timeout_DNS" error in retry rules, also documented since at least
6702 1.62, also never worked. As it isn't clear exactly what this means, and
6703 clearly it isn't a major issue, I have abolished the feature by treating it
6704 as "timeout", and writing a warning to the main and panic logs.
6706 36. The display of retry rules for -brt wasn't always showing the error code
6709 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6710 timeout_connect_A, timeout_connect_MX.
6712 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6713 to the empty sender.
6715 39. The daemon was not analysing the content of -oX till after it had closed
6716 stderr and disconnected from the controlling terminal. This meant that any
6717 syntax errors were only noted on the panic log, and the return code from
6718 the command was 0. By re-arranging the code a little, I've made the
6719 decoding happen first, so such errors now appear on stderr, and the return
6720 code is 1. However, the actual setting up of the sockets still happens in
6721 the disconnected process, so errors there are still only recorded on the
6724 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6725 connections (as happens on some IP stacks) was logged at start up time as
6726 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6727 it from "IPv6 and IPv4", which means that two separate sockets are being
6730 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6731 says whether AF_INET or AF_INET6 was passed as an argument.
6733 42. Exiwhat output was messed up when time zones were included in log
6736 43. Exiwhat now gives more information about the daemon's listening ports,
6737 and whether -tls-on-connect was used.
6739 44. The "port" option of the smtp transport is now expanded.
6741 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6742 silently ignored. Now an error message is written to the main and panic
6745 46. There's a new ACL modifier called "logwrite" which writes to a log file
6746 as soon as it is encountered.
6748 47. Added $local_user_uid and $local_user_gid at routing time.
6750 48. Exim crashed when trying to verify a sender address that was being
6753 49. Exim was recognizing only a space character after ".include". It now also
6754 recognizes a tab character.
6756 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6757 extracting the relevant information from the specification. The man page no
6758 longer contains scrambled data for the -d option, and I've added a section
6759 at the front about calling Exim under different names.
6761 51. Added "extra_headers" argument to the "mail" command in filter files.
6763 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6766 53. Installed eximstats 1.29.
6768 54. Added transport_filter_timeout as a generic transport option.
6770 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6771 Cc: header lines. This was required by RFC 822, but it not required by RFC
6774 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6775 incoming messages that did not have them. Now it does so only if the
6776 message originates locally, that is, if there is no associated remote host
6777 address. When Resent- header lines are present, this applies to the Resent-
6778 lines rather than the non-Resent- lines.
6780 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6781 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6783 58. Messages for configuration errors now include the name of the main
6784 configuration file - useful now that there may be more than one file in a
6785 list (.included file names were always shown).
6787 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6788 for those rare installations that do not start the daemon as root or run it
6789 setuid root. I've cut out the call to initgroups() if the daemon is not
6792 60. The Exim user and group can now be bound into the binary as text strings
6793 that are looked up at the start of Exim's processing.
6795 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6797 62. Added $mailstore_basename variable.
6799 63. Installed patch to sieve.c from Michael Haardt.
6801 64. When Exim failed to open the panic log after failing to open the main log,
6802 the original message it was trying to log was written to stderr and debug
6803 output, but if they were not available (the usual case in production), it
6804 was lost. Now it is written to syslog before the two lines that record the
6805 failures to open the logs.
6807 65. Users' Exim filters run in subprocesses under the user's uid. It is
6808 possible for a "deliver" command or an alias in a "personal" command to
6809 provoke an address rewrite. If logging of address rewriting is configured,
6810 this fails because the process is not running as root or exim. There may be
6811 a better way of dealing with this, but for the moment (because 4.30 needs
6812 to be released), I have disabled address rewrite logging when running a
6813 filter in a non-root, non-exim process.
6819 1. The buildconfig auxiliary program wasn't quoting the value set for
6820 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6821 not defined. This bug was masked in 4.22 by the effect that was fixed in
6824 2. Some messages that were rejected after a message id was allocated were
6825 shown as "incomplete" by exigrep. It no longer does this for messages that
6826 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6828 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6829 have allow_domain_literals set, the ID did not get logged in the <= line.
6830 Domain literals are now always recognized in Message-ID: header lines.
6832 4. The first argument for a ${extract expansion item is the key name or field
6833 number. Leading and trailing spaces in this item were not being ignored,
6834 causing some misleading effects.
6836 5. When deliver_drop_privilege was set, single queue runner processes started
6837 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6838 same command in the process it spins off) were not dropping privilege.
6840 6. When the daemon running as "exim" started a queue runner, it always
6841 re-executed Exim in the spun-off process. This is a waste of effort when
6842 deliver_drop_privilege is set. The new process now just calls the
6843 queue-runner function directly.
6849 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6852 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6853 ignored. Though the use of -oP was forcing the writing of a pid file, it
6854 was always written to the default place.
6856 3. If the message "no IP address found for host xxxx" is generated during
6857 incoming verification, it is now followed by identification of the incoming
6858 connection (so you can more easily find what provoked it).
6860 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6862 5. Added some features to "harden" Exim a bit more against certain attacks:
6864 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6865 be put in Local/Makefile. This is like the never_users runtime option,
6866 but it cannot be overridden. The default setting is "root".
6868 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6869 prefix string with which any file named in a -C command line option
6872 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6873 is retained for -C and -D only if the caller of Exim is root. Without
6874 it, the exim user may also use -C and -D and retain privilege.
6876 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6877 command line option is disabled.
6879 6. Macro names set by the -D option must start with an upper case letter, just
6880 like macro names defined in the configuration file.
6882 7. Added "dereference=" facility to LDAP.
6884 8. Two instances of the typo "uknown" in the source files are fixed.
6886 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6887 the Configure-Makefile script screwed up while processing it.
6889 10. Incorporated PCRE 4.4.
6891 11. The SMTP synchronization check was not operating right at the start of an
6892 SMTP session. For example, it could not catch a HELO sent before the client
6893 waited for the greeting. There is now a check for outstanding input at the
6894 point when the greeting is written. Because of the duplex, asynchronous
6895 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6896 way, but not yet received, when the check is performed.
6898 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6899 on TCP/IP sockets, because this apparently causes some broken clients to
6902 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6903 unchanged) from the Cygwin maintainer.
6905 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6906 was supported instead of testing for mbx. Effectively a typo.
6908 15. The spa authenticator server code was not checking that the input it
6909 received was valid base64.
6911 16. The debug output line for the "set" modifier in ACLs was not showing the
6912 name of the variable that was being set.
6914 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6916 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6917 lookup on its own. Something else was needed to trigger the lookup. For
6918 example, a match in host_lookup or the need for a host name in a host list.
6919 Now, if $sender_host_name is referenced and the host name has not yet been
6920 looked up, a lookup is performed. If the lookup fails, the variable remains
6921 empty, and $host_lookup_failed is set to "1".
6923 19. Added "eqi" as a case-independent comparison operator.
6925 20. The saslauthd authentication condition could segfault if neither service
6926 nor realm was specified.
6928 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6929 error message that was logged was misleading, and incoming SMTP
6930 connections were dropped. The message is now more accurate, and temporary
6931 errors are given to SMTP connections.
6933 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6934 (except RSET) with a 421 error, until QUIT is received. However, it was
6935 failing to send a response to QUIT.
6937 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6938 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6939 ACL is now run after the helo_try_verify_hosts code.
6941 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6942 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6943 case-independent, so other case variants are also recognized). Apparently
6944 some systems use these upper case variants.
6946 25. If more than two messages were waiting for the same host, and a transport
6947 filter was specified for the transport, Exim sent two messages over the
6948 same TCP/IP connection, and then failed with "socket operation on non-
6949 socket" when it tried to send the third.
6951 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6953 27. The extern definition of crypt16() in expand.c was not being excluded when
6954 the OS had its own crypt16() function.
6956 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6957 as a preferred synonym for return_size_limit, both as an option and as an
6960 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6962 30. Changed the default configuration ACL to relax the local part checking rule
6963 for addresses that are not in any local domains. For these addresses,
6964 slashes and pipe symbols are allowed within local parts, but the sequence
6965 /../ is explicitly forbidden.
6967 31. SPA server authentication was not clearing the challenge buffer before
6970 32. log_message in a "warn" ACL statement was writing to the reject log as
6971 well as to the main log, which contradicts the documentation and doesn't
6972 seem right (because no rejection is happening). So I have stopped it.
6974 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6975 However, I am unable to do any testing of this.
6977 34. Fixed an infelicity in the appendfile transport. When checking directories
6978 for a mailbox, to see if any needed to be created, it was accidentally
6979 using path names with one or more superfluous leading slashes; tracing
6980 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6982 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6983 contents are added to the log line that is written for every discarded
6984 recipient. (Previously a log_message setting was ignored.)
6986 36. The ${quote: operator now quotes the string if it is empty.
6988 37. The install script runs exim in order to find its version number. If for
6989 some reason other than non-existence or emptiness, which it checks, it
6990 could not run './exim', it was installing it with an empty version number,
6991 i.e. as "exim-". This error state is now caught, and the installation is
6994 38. An argument was missing from the function that creates an error message
6995 when Exim fails to connect to the socket for saslauthd authentication.
6996 This could cause Exim to crash, or give a corrupted message.
6998 39. Added isip, isip4, and isip6 to ${if conditions.
7000 40. The ACL variables $acl_xx are now saved with the message, and can be
7001 accessed later in routers, transports, and filters.
7003 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7004 strings in the file are not string-expanded.
7006 42. If a MAIL command specified a SIZE value that was too large to fit into an
7007 int variable, the check against message_size_limit failed. Such values are
7008 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7009 day this will have to be increased, but I don't think I want to be around
7010 when emails are that large.
7017 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7018 iconv() is not standard in FreeBSD.
7020 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7021 IPv6 enabled. The observed symptom was a segmentation fault on return from
7022 the function os_common_find_running_interfaces() in src/os.c.
7024 3. In the check_special_case() function in daemon.c I had used "errno" as an
7025 argument name, which causes warnings on some systems. This was basically a
7026 typo, since it was named "eno" in the comments!
7028 4. The code that waits for the clock to tick (at a resolution of some fraction
7029 of a second) so as to ensure message-id uniqueness was always waiting for
7030 at least one whole tick, when it could have waited for less. [This is
7031 almost certainly not relevant at current processor speeds, where it is
7032 unlikely to ever wait at all. But we try to future-proof.]
7034 5. The function that sleeps for a time interval that includes fractions of a
7035 second contained a race. It did not block SIGALRM between setting the
7036 timer, and suspending (a couple of lines later). If the interval was short
7037 and the sigsuspend() was delayed until after it had expired, the suspension
7038 never ended. On busy systems this could lead to processes getting stuck for
7041 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7042 process, before it forks any delivery processes. The open lookup caching
7043 mechanism meant that the open file or database connection was passed into
7044 the delivery process. The problem was that delivery processes always tidy
7045 up cached lookup data. This could cause a problem for the next delivery
7046 process started by the queue runner, because the external queue runner
7047 process does not know about the closure. So the next delivery process
7048 still has data in the lookup cache. In the case of a file lookup, there was
7049 no problem because closing a file descriptor in a subprocess doesn't affect
7050 the parent. However, if the lookup was caching a connection to a database,
7051 the connection was closed, and the second delivery process was likely to
7052 see errors such as "PGSQL: query failed: server closed the connection
7053 unexpectedly". The problem has been fixed by closing all cached lookups
7054 in a queue runner before running a delivery process.
7056 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7057 seem to have the "const" qualifier which it has on other OS. I've
7060 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7061 *of the same type* (client or server) with the same public name that an
7062 error should be diagnosed.
7064 9. When Exim looked up a host name for an IP address, but failed to find the
7065 original IP address when looking up the host name (a safety check), it
7066 output the message "<ip address> does not match any IP for NULL", which was
7067 confusing, to say the least. The bug was that the host name should have
7068 appeared instead of "NULL".
7070 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7071 user that is built into the binary, and the -C or -D options is used, root
7072 privilege is dropped before the configuration file is read. In addition,
7073 logging is switched to stderr instead of the normal log files. If the
7074 configuration then re-defines the Exim user, the unprivileged environment
7075 is probably not what is expected, so Exim logs a panic warning message (but
7078 However, if deliver_drop_privilege is set, the unprivileged state may well
7079 be exactly what is intended, so the warning has been cut out in that case,
7080 and Exim is allowed to try to write to its normal log files.
7086 1. smtp_return_error_details was not giving details for temporary sender
7087 or receiver verification errors.
7089 2. Diagnose a configuration error if two authenticators have the same public
7092 3. Exim used not to create the message log file for a message until the first
7093 delivery attempt. This could be confusing when incoming messages were held
7094 for policy or load reasons. The message log file is now created at the time
7095 the message is received, and an initial "Received" line is written to it.
7097 4. The automatically generated man page for command line options had a minor
7098 bug that caused no ill effects; however, a more serious problem was that
7099 the procedure for building the man page automatically didn't always
7100 operate. Consequently, release 4.20 contains an out-of-date version. This
7101 shouldn't happen again.
7103 5. When building Exim with embedded Perl support, the script that builds the
7104 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7105 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7107 6. The freeze_tell option was not being used for messages that were frozen on
7108 arrival, either by an ACL or by local_scan().
7110 7. Added the smtp_incomplete_transaction log selector.
7112 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7113 was accepting AUTH without a new EHLO.
7115 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7116 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7119 10. Logging of TCP/IP connections (when configured) now happens in the main
7120 daemon process instead of the child process, so that the TCP/IP connection
7121 count is more accurate (but it can never be perfect).
7123 11. The use of "drop" in a nested ACL was not being handled correctly in the
7124 outer ACL. Now, if condition failure induced by the nested "drop" causes
7125 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7126 or "require"), the connection is dropped.
7128 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7129 that yield "discard" can now be used with an "accept" or a "discard" verb,
7130 but an error is generated for any others (because I can't see a useful way
7131 to define what should happen).
7133 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7134 are now processed in the same way as lines in the Exim configuration file.
7135 In particular, continuation lines are supported.
7137 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7139 15. Added -ti meaning -t -i.
7141 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7142 domains, and warn by logging if others are found.
7144 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7145 returned from GIFCONF when Exim is trying to find the list of interfaces on
7146 a host. The code in os.c has been modified to copy each ifreq to an aligned
7147 structure in all cases.
7149 Also, in some cases, the returned ifreq's were being copied to a 'struct
7150 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7151 means the last couple of bytes of an IPv6 address could be chopped if the
7152 ifreq contained only a normal sockaddr (14 bytes storage).
7154 18. Named domain lists were not supported in the hosts_treat_as_local option.
7155 An entry such as +xxxx was not recognized, and was treated as a literal
7158 19. Ensure that header lines added by a DATA ACL are included in the reject log
7159 if the ACL subsequently rejects the message.
7161 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7162 MD5 (which is deprecated).
7164 21. When testing a filter file using -bf, Exim was writing a message when it
7165 took the sender from a "From " line in the message, but it was not doing so
7166 when it took $return_path from a Return-Path: header line. It now does.
7168 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7169 with a valid header line field name (a series of printing characters
7170 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7172 23. Changed "disc" in the source to "disk" to conform to the documentation and
7173 the book and for uniformity.
7175 24. Ignore Sendmail's -Ooption=value command line item.
7177 25. When execve() failed while trying to run a command in a pipe transport,
7178 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7179 could be confused with a return value of 69 from the command itself. This
7180 has been changed to 127, the value the shell returns if it is asked to run
7181 a non-existent command. The wording for the related log line suggests a
7182 non-existent command as the problem.
7184 26. If received_header_text expands to an empty string, do not add a Received:
7185 header line to the message. (Well, it adds a token one on the spool, but
7186 marks it "old" so that it doesn't get used or transmitted.)
7188 27. Installed eximstats 1.28 (addition of -nt option).
7190 28. There was no check for failure on the call to getsockname() in the daemon
7191 code. This can fail if there is a shortage of resources on the system, with
7192 ENOMEM, for example. A temporary error is now given on failure.
7194 29. Contrary to the C standard, it seems that in some environments, the
7195 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7196 program. Exim now does this explicitly; it affects the formatting of
7197 timestamps using strftime().
7199 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7200 complaints. I've now initialized all the variables, to avoid this.
7202 32. Header lines added by a system filter were not being "seen" during
7203 transport-time rewrites.
7205 33. The info_callback() function passed to OpenSSL is set up with type void
7206 (*)(SSL *, int, int), as described somewhere. However, when calling the
7207 function (actually a macro) that sets it up, the type void(*)() is
7208 expected. I've put in a cast to prevent warnings from picky compilers.
7210 34. If a DNS black list lookup found a CNAME record, but there were no A
7211 records associated with the domain it pointed at, Exim crashed.
7213 35. If a DNS black list lookup returned more than one A record, Exim ignored
7214 all but the first. It now scans all returned addresses if a particular IP
7215 value is being sought. In this situation, the contents of the
7216 $dnslist_value variable are a list of all the addresses, separated by a
7219 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7220 to accept a host name and all its aliases if the forward lookup for any of
7221 them yielded the IP address of the incoming connection. Now it accepts only
7222 those names whose forward lookup yields the correct IP address. Any other
7223 names are discarded. This closes a loophole whereby a rogue DNS
7224 administrator could create reverse DNS records to break through a
7225 wildcarded host restriction in an ACL.
7227 37. If a user filter or a system filter that ran in a subprocess used any of
7228 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7229 the wrong values were passed to the pipe command ($thisaddress had the
7230 value of $0, $0 had the value of $1, etc). This bug was introduced by
7231 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7233 38. Improved the line breaking for long SMTP error messages from ACLs.
7234 Previously, if there was no break point between 40 and 75 characters, Exim
7235 left the rest of the message alone. Two changes have been made: (a) I've
7236 reduced the minimum length to 35 characters; (b) if it can't find a break
7237 point between 35 and 75 characters, it looks ahead and uses the first one
7238 that it finds. This may give the occasional overlong line, but at least the
7239 remaining text gets split now.
7241 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7242 file descriptors might be low, and that setting 1000 would always raise it.
7243 It turns out that in some environments, the limit is already over 1000 and
7244 that lowering it causes trouble. So now Exim takes care not to decrease it.
7246 40. When delivering a message, the value of $return_path is set to $sender_
7247 address at the start of routing (routers may change the value). By an
7248 oversight, this default was not being set up when an address was tested by
7249 -bt or -bv, which affected the outcome if any router or filter referred to
7252 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7253 log only when "message" or "log_message" are set. However, if one of the
7254 conditions was an address verification, or a call to a nested ACL, the
7255 messages generated by the underlying test were being passed through. This
7256 no longer happens. The underlying message is available in $acl_verify_
7257 message for both "message" and "log_message" expansions, so it can be
7258 passed through if needed.
7260 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7261 new expansion $bh_ to give the encoded byte string without charset
7262 translation. Translation happens only if iconv() is available; HAVE_ICONV
7263 indicates this at build time. HEADERS_CHARSET gives the charset to
7264 translate to; headers_charset can change it in the configuration, and
7265 "headers charset" can change it in an individual filter file.
7267 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7268 that creates RFC 2047 encoded "words" labels them as that charset instead
7269 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7270 expansion operator; (ii) when Exim creates a From: line for a local
7271 message; (iii) when a header line is rewritten to include a "phrase" part.
7273 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7274 buggy, causing it to skip the first lines of messages whose message ID
7275 ended in 'D'. This would not have bitten before Exim release 4.14, because
7276 message IDs were unlikely to end in 'D' before then. The effect was to have
7277 incorrect size information for certain domains.
7279 45. #include "config.h" was missing at the start of the crypt16.c module. This
7280 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7283 46. If there was a timeout during a "random" callout check, Exim treated it as
7284 a failure of the random address, and carried on sending RSET and the real
7285 address. If the delay was just some slowness somewhere, the response to the
7286 original RCPT would be taken as a response to RSET and so on, causing
7287 mayhem of various kinds.
7289 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7290 when I implemented it. It didn't allow for the fact that some option values
7291 may legitimately be negative (e.g. size_addition), and it didn't even do
7292 the right test for positive values.
7294 48. Domain names in DNS records are case-independent. Exim always looks them up
7295 in lower case. Some resolvers return domain names in exactly the case they
7296 appear in the zone file, that is, they may contain uppercase letters. Not
7297 all resolvers do this - some return always lower case. Exim was treating a
7298 change of case by a resolver as a change of domain, similar to a widening
7299 of a domain abbreviation. This triggered its re-routing code and so it was
7300 trying to route what was effectively the same domain again. This normally
7301 caused routing to fail (because the router wouldn't handle the domain
7302 twice). Now Exim checks for this case specially, and just changes the
7303 casing of the domain that it ultimately uses when it transmits the message
7306 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7309 50. If a filter generated a file delivery with a non-absolute name (possible if
7310 no home directory exists for the router), the forbid_file option was not
7313 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7314 the existing equality matching.
7316 52. Exim was using ints instead of ino_t variables in some places where it was
7317 dealing with inode numbers.
7319 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7320 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7321 called TMPDIR, and if it finds it is different, it changes its value.
7323 54. The smtp_printf() function is now made available to local_scan() so
7324 additional output lines can be written before returning. There is also an
7325 smtp_fflush() function to enable the detection of a dropped connection.
7326 The variables smtp_input and smtp_batched_input are exported to
7329 55. Changed the default runtime configuration: the message "Unknown user"
7330 has been removed from the ACL, and instead placed on the localuser router,
7331 using the cannot_route_message feature. This means that any verification
7332 failures that generate their own messages won't get overridden. Similarly,
7333 the "Unrouteable address" message that was in the ACL for unverifiable
7334 relay addresses has also been removed.
7336 56. Added hosts_avoid_esmtp to the smtp transport.
7338 57. The exicyclog script was not checking for the esoteric option
7339 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7340 will work only if exicyclog is run under the appropriate euid.
7342 58. Following a discussion on the list, the rules by which Exim recognises line
7343 endings on incoming messages have been changed. The -dropcr and drop_cr
7344 options are now no-ops, retained only for backwards compatibility. The
7345 following line terminators are recognized: LF CRLF CR. However, special
7346 processing applies to CR:
7348 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7349 nor a local message in the state where . is a terminator.
7351 (ii) If a bare CR is encountered in a header line, an extra space is added
7352 after the line terminator so as not to end the header. The reasoning
7353 behind this is that bare CRs in header lines are most likely either
7354 to be mistakes, or people trying to play silly games.
7356 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7357 was being incorrectly given as 18 bytes larger than it should have been.
7358 This is a VOB (very old bug).
7360 60. This may never have affected anything current, but just in case it has:
7361 When the local host is found other than at the start of a list of hosts,
7362 the local host, those with the same MX, and any that follow, are discarded.
7363 When the list in question was part of a longer list of hosts, the following
7364 hosts (not currently being processed) were also being discarded. This no
7365 longer happens. I'm not sure if this situation could ever has previously
7368 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7371 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7372 header. This is something that is recommended in a new Internet Draft, and
7373 is something that is documented as being done by Sendmail. There are two
7374 possible values. For messages generated by the autoreply transport, Exim
7377 Auto-Submitted: auto-replied
7379 whereas for all other generated messages (e.g. bounces) it adds
7381 Auto-Submitted: auto-generated
7383 63. The "personal" condition in filters now includes a test for the
7384 Auto-Submitted: header. If it contains the string "auto-" the message it
7385 not considered personal.
7387 64. Added rcpt_include_affixes as a generic transport option.
7389 65. Added queue_only_override (default true).
7391 66. Added the syslog_duplication option.
7393 67. If what should have been the first header line of a message consisted of
7394 a space followed by a colon, Exim was mis-interpreting it as a header line.
7395 It isn't of course - it is syntactically invalid and should therefore be
7396 treated as the start of the message body. The misbehaviour could have
7397 caused a number of strange effects, including loss of data in subsequent
7398 header lines, and spool format errors.
7400 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7401 client host had authenticated. This control can now be exercised by an ACL
7402 for more flexibility.
7404 69. By default, callouts do not happen when testing with -bh. There is now a
7405 variant, -bhc, which does actually run the callout code, including
7406 consulting and updating the callout cache.
7408 70. Added support for saslauthd authentication, courtesy of Alexander
7411 71. If statvfs() failed on the spool or log directories while checking their
7412 size for availability, Exim confusingly gave the error "space shortage".
7413 Furthermore, in debugging mode it crashed with a floating point exception.
7414 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7415 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7416 serious problem, Exim now writes to the main and panic logs when this
7417 happens, with details of the failure. It then refuses to accept the
7418 incoming message, giving the message "spool directory problem" or "log
7419 directory problem" with a 421 code for SMTP messages.
7421 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7422 0, 1, and 2 exist, because some OS complain for execs without them (see
7423 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7424 descriptors. However, the code omitted to check that the open succeeded,
7425 causing mysterious errors if for some reason the permissions on /dev/null
7426 got screwed. Now Exim writes a message to the main and panic logs, and
7427 bombs out if it can't open /dev/null.
7429 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7430 interact so that it is all more flexible. It is supposed to remain
7431 backwards compatible. Also added extra_local_interfaces.
7433 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7434 to bomb out with an assertion failure - to the client this appears as a
7435 connection drop. This problem occurs in the part of the code that was taken
7436 from the Samba project. Fortunately, the assertion is in a very simple
7437 function, so I have fixed this by reproducing the function inline in the
7438 one place where it is called, and arranging for authentication to fail
7439 instead of killing the process with assert().
7441 75. The SPA client code was not working when the server requested OEM rather
7442 than Unicode encoding.
7444 76. Added code to make require_files with a specific uid setting more usable in
7445 the case where statting the file as root fails - usually a non-root-mounted
7446 NFS file system. When this happens and the failure is EACCES, Exim now
7447 forks a subprocess and does the per-uid checking as the relevant uid.
7449 77. Added process_log_path.
7451 78. If log_file_path was not explicitly set, a setting of check_log_space or
7452 check_log_inodes was ignored.
7454 79. If a space check for the spool or log partitions fails, the incident is now
7455 logged. Of course, in the latter case the data may get lost...
7457 80. Added the %p formatting code to string_format() so that it can be used to
7458 print addresses in debug_print(). Adjusted all the address printing in the
7459 debugging in store.c to use %p rather than %d.
7461 81. There was a concern that a line of code in smtp_in.c could overflow a
7462 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7463 initially expressed, the concern was not well-founded, because trailing
7464 spaces are removed early. However, if the trailing spaces were followed by
7465 a NULL, they did not get removed, so the overflow was possible. Two fixes
7468 (a) I re-wrote the offending code in a cleaner fashion.
7469 (b) If an incoming SMTP command contains a NULL character, it is rejected
7472 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7473 runs initgroups(), so that if the Exim user is in any additional groups,
7474 they will be used during message reception.
7480 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob