1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platform supports and has the facility enabled, it will
125 be requested on all coneections.
127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
130 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
132 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
133 success-DSN messages. Previously the From: header was always the default
134 one for these; the option was ignored.
136 JH/28 Fix the timeout on smtp response to apply to the whole response.
137 Previously it was reset for every read, so a teergrubing peer sending
138 single bytes within the time limit could extend the connection for a
139 long time. Credit to Qualsys Security Advisory Team for the discovery.
141 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
142 delivery address, which leaked information of the results of local
143 forwarding. Change to the original envelope recipient address, per
146 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
147 requested. Previously not bounce was generated and a log entry of
148 error ignored was made.
150 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
152 JH/32 Introduce a general tainting mechanism for values read from the input
153 channel, and values derived from them. Refuse to expand any tainted
154 values, to catch one form of exploit.
156 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
157 was unused and the unexpanded text used for the test. Found and
158 fixed by Ruben Jenster.
164 JH/01 Remove code calling the customisable local_scan function, unless a new
165 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
167 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
168 non-signal-safe functions being used.
170 JH/03 Bug 2269: When presented with a received message having a stupidly large
171 number of DKIM-Signature headers, disable DKIM verification to avoid
172 a resource-consumption attack. The limit is set at twenty.
174 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
175 report of oldest_pass in ${authres } in consequence, and separate out
176 some descriptions of reasons for verification fail.
178 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
179 files in the spool were present and unlocked. A queue-runner could spot
180 them, resulting in a duplicate delivery. Fix that by doing the unlock
181 after the unlink. Investigation by Tim Stewart. Take the opportunity to
182 add more error-checking on spoolfile handling while that code is being
185 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
186 No known attacks, no CVE, this is defensive hardening.
188 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
189 a queue-runner could start a delivery while other operations were ongoing.
190 Cutthrough delivery was a common victim, resulting in duplicate delivery.
191 Found and investigated by Tim Stewart. Fix by using the open message data
192 file handle rather than opening another, and not locally closing it (which
193 releases a lock) for that case, while creating the temporary .eml format
194 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
196 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
197 $sender_verify_failure/$recipient_verify_failure to "random".
199 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
202 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
203 Previously this would segfault.
205 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
208 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
209 like zero, since the resolver should be doing this for us, But we need one
210 as a CNAME but no MX presence gets the CNAME returned; we need to check
211 that doesn't point to an MX to declare it "no MX returned" rather than
212 "error, loop". A new main option is added so the older capability of
213 following some limited number of chain links is maintained.
215 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
217 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
218 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
219 now-deprecated earlier definitions used only specified the range up to TLS
220 1.2 (in the older-version library docs).
222 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
224 JH/15 Rework TLS client-side context management. Stop using a global, and
225 explicitly pass a context around. This enables future use of TLS for
226 connections to service-daemons (eg. malware scanning) while a client smtp
227 connection is using TLS; with cutthrough connections this is quite likely.
229 JH/16 Fix ARC verification to do AS checks in reverse order.
231 JH/17 Support a "tls" option on the ${readsocket } expansion item.
233 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
234 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
235 Previously the "utf8" would be re-prepended for every additional message.
237 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
238 Previously thery were accepted, resulting in issues when attempting to
239 forward messages to a non-supporting MTA.
241 PP/02 Let -n work with printing macros too, not just options.
243 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
244 one parent address was copied, and bogus data was used at delivery-logging
245 time. Either a crash (after delivery) or bogus log data could result.
246 Discovery and analysis by Tim Stewart.
248 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
249 Previously if the string ended mid-character, we did not insert the
250 promised '?' replacement.
252 PP/04 Documentation: current string operators work on bytes, not codepoints.
254 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
255 should pack well giving a smaller memory footprint so better caching and
256 therefore performance. Group the declarations where this can't be done so
257 that the byte-sized flag variables are not interspersed among pointer
258 variables, giving a better chance of good packing by the compiler.
260 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
261 non-null, to avoid issues with sites running BATV. Previously reports were
262 sent with an empty envelope sender so looked like bounces.
264 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
265 The ignore_error flag wasn't being returned from the filter subprocess so
266 was not set for later routers. Investigation and fix by Matthias Kurz.
268 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
269 and a msg:complete for the whole, when a message is manually removed using
270 -Mrm. Developement by Matthias Kurz, hacked on by JH.
272 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
273 a "Gnu special" function, asprintf() in the DB utility binary builds; I
274 hope that is portable enough.
276 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
277 requiring a known-CA anchor certificate; make it now rely entirely on the
278 TLSA as an anchor. Checking the name on the leaf cert against the name
279 on the A-record for the host is still done for TA (but not for EE mode).
281 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
282 included in delivery lines for non-proxied connections, when compiled with
283 SUPPORT_SOCKS and running with proxy logging enabled.
285 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
286 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
287 move the existing event to fire before the normal logging of message
288 failure so that custom logging is bracketed by normal logging.
290 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
291 msg:fail:internal event. Developement by Matthias Kurz.
293 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
294 far too small for todays use of crypto signatures stored there. Go all
295 the way to the max DNS message size of 64kB, even though this might be
296 overmuch for IOT constrained device use.
298 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
299 copy a string over itself. The library routine is documented as not
300 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
302 JH/32 For main options check_spool_space and check_inode_space, where the
303 platform supports 64b integers, support more than the previous 2^31 kB
304 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
305 the previous G, M, k.
307 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
308 $authenticated_fail_id variable on authentication failure. Previously
311 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
312 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
313 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
314 GNUTLS_SEC_PARAM_MEDIUM.
316 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
317 side. Previously we would continue as if no SNI had been received.
319 JH/36 Harden the handling of string-lists. When a list consisted of a sole
320 "<" character, which should be a list-separator specification, we walked
321 off past the nul-terimation.
323 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
324 causes) even when the retry time is not yet met. Previously they were
325 not, meaning that when (say) an account was over-quota and temp-rejecting,
326 and multiple senders' messages were queued, only one sender would get
327 notified on each configured delay_warning cycle.
329 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
331 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
332 cipher-suites, an error can be left on the stack even for a succeeding
333 accept; this results in impossible error messages when a later operation
336 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
337 return error codes indicating retry. Under TLS1.3 this becomes required.
339 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
340 it only wrote the new authenticators, resulting in a lack of tracking of
341 peer changes of ESMTP extensions until the next cache flush.
343 JH/41 Fix the loop reading a message header line to check for integer overflow,
344 and more-often against header_maxsize. Previously a crafted message could
345 induce a crash of the recive process; now the message is cleanly rejected.
347 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
348 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
354 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
355 When redis_servers is set to a list of > 1 element, and the Redis servers
356 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
357 case of MOVED into a DEFER case instead, thus moving the query onto the
358 next server in the list. For a cluster of N elements, all N servers must
359 be defined in redis_servers.
361 GF/02 Catch and remove uninitialized value warning in exiqsumm
362 Check for existence of @ARGV before looking at $ARGV[0]
364 JH/01 Replace the store_release() internal interface with store_newblock(),
365 which internalises the check required to safely use the old one, plus
366 the allocate and data copy operations duplicated in both (!) of the
367 extant use locations.
369 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
370 modifier. This matches the restriction on the commandline.
372 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
373 Previously only the last row was returned.
375 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
376 we assumed that tags in the header were well-formed, and parsed the
377 element content after inspecting only the first char of the tag.
378 Assumptions at that stage could crash the receive process on malformed
381 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
382 While running the DKIM ACL we operate on the Permanent memory pool so that
383 variables created with "set" persist to the DATA ACL. Also (at any time)
384 DNS lookups that fail create cache records using the Permanent pool. But
385 expansions release any allocations made on the current pool - so a dnsdb
386 lookup expansion done in the DKIM ACL releases the memory used for the
387 DNS negative-cache, and bad things result. Solution is to switch to the
388 Main pool for expansions.
389 While we're in that code, add checks on the DNS cache during store_reset,
390 active in the testsuite.
391 Problem spotted, and debugging aided, by Wolfgang Breyha.
393 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
394 When none of the hosts presented to a transport match an already-open
395 connection, close it and proceed with the list. Previously we would
396 queue the message. Spotted by Lena with Yahoo, probably involving
399 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
400 Previously a spurious "250 OK id=" response was appended to the proper
403 JH/08 The "support for" informational output now, which built with Content
404 Scanning support, has a line for the malware scanner interfaces compiled
405 in. Interface can be individually included or not at build time.
407 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
408 by the template makefile "src/EDITME". The "STREAM" support for an older
409 ClamAV interface method is removed.
411 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
412 rows affected is given instead).
414 JH/11 The runtime Berkeley DB library version is now additionally output by
415 "exim -d -bV". Previously only the compile-time version was shown.
417 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
418 SMTP connection. Previously, when one had more receipients than the
419 first, an abortive onward connection was made. Move to full support for
420 multiple onward connections in sequence, handling cutthrough connection
421 for all multi-message initiating connections.
423 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
424 routers. Previously, a multi-recipient message would fail to match the
425 onward-connection opened for the first recipient, and cause its closure.
427 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
428 a timeout on read on a GnuTLS initiating connection, resulting in the
429 initiating connection being dropped. This mattered most when the callout
430 was marked defer_ok. Fix to keep the two timeout-detection methods
433 JH/15 Relax results from ACL control request to enable cutthrough, in
434 unsupported situations, from error to silently (except under debug)
435 ignoring. This covers use with PRDR, frozen messages, queue-only and
438 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
440 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
441 metadata, resulting in a crash in free().
443 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
444 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
445 Broken also in d185889f4, with init system revamp.
447 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
448 Previously we abruptly closed the connection after reading a malware-
449 found indication; now we go on to read the "scan ok" response line,
452 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
453 ACL. Previously, a crash would result.
455 JH/19 Speed up macro lookups during configuration file read, by skipping non-
456 macro text after a replacement (previously it was only once per line) and
457 by skipping builtin macros when searching for an uppercase lead character.
459 JH/20 DANE support moved from Experimental to mainline. The Makefile control
460 for the build is renamed.
462 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
463 was allocated for every new TLS startup, meaning one per message. Fix
464 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
466 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
467 reported the original. Fix to report (as far as possible) the ACL
468 result replacing the original.
470 JH/23 Fix memory leak during multi-message connections using STARTTLS under
471 OpenSSL. Certificate information is loaded for every new TLS startup,
472 and the resources needed to be freed.
474 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
476 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
479 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
480 DATA response info to the (existing) per-recipient response info for
481 the "C=" log element. It can have useful tracking info from the
482 destination system. Patch from Simon Arlott.
484 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
485 length value. Previously this would segfault.
487 HS/02 Support Avast multiline protoocol, this allows passing flags to
488 newer versions of the scanner.
490 JH/28 Ensure that variables possibly set during message acceptance are marked
491 dead before release of memory in the daemon loop. This stops complaints
492 about them when the debug_store option is enabled. Discovered specifically
493 for sender_rate_period, but applies to a whole set of variables.
494 Do the same for the queue-runner and queue-list loops, for variables set
495 from spool message files. Do the same for the SMTP per-message loop, for
496 certain variables indirectly set in ACL operations.
498 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
499 as a multi-recipient message from a mailinglist manager). The coding had
500 an arbitrary cutoff number of characters while checking for more input;
501 enforced by writing a NUL into the buffer. This corrupted long / fast
502 input. The problem was exposed more widely when more pipelineing of SMTP
503 responses was introduced, and one Exim system was feeding another.
504 The symptom is log complaints of SMTP syntax error (NUL chars) on the
505 receiving system, and refused recipients seen by the sending system
506 (propating to people being dropped from mailing lists).
507 Discovered and pinpointed by David Carter.
509 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
510 replaced by the ${authresults } expansion.
512 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
514 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
515 allows proper process termination in container environments.
517 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
518 Previously the "final dot" had a newline after it; ensure it is CR,LF.
520 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
521 and "err_perm", deprecated since 4.83 when the RFC-defined words
522 "temperror" and "permerror" were introduced.
524 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
525 transport-filters or DKIM-signing. The restriction was lost in the
526 consolidation of verify-callout and delivery SMTP handling.
527 Extend the restriction to also cover ARC-signing.
529 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
530 in defer=pass mode supply a 450 to the initiator. Previously the message
533 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
534 tls_require_ciphers is used as before.
536 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
537 "pass_unscanned". Only tmpfails from the scanner are written to
538 the paniclog, as they may require admin intervention (permission
539 denied, license issues). Other scanner errors (like decompression
540 bombs) do not cause a paniclog entry.
542 JH/36 Fix reinitialisation of DKIM logging variable between messages.
543 Previously it was possible to log spurious information in receive log
546 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
547 triggered odd behaviour from Outlook Express clients.
549 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
552 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
553 since the IETF WG has not yet settled on that versus the original
554 "bare" representation.
556 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
557 Previously the millisecond value corrupted the output.
558 Fix also for syslog_pid=no and log_selector +pid, for which the pid
559 corrupted the output.
565 JH/01 Rework error string handling in TLS interface so that the caller in
566 more cases is responsible for logging. This permits library-sourced
567 string to be attached to addresses during delivery, and collapses
568 pairs of long lines into single ones.
570 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
571 during configuration. Wildcards are allowed and expanded.
573 JH/02 Rework error string handling in DKIM to pass more info back to callers.
574 This permits better logging.
576 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
577 do not close it down and have the child transport start it up again on
578 the passed-on TCP connection. Instead, proxy the child (and any
579 subsequent ones) for TLS via a unix-domain socket channel. Logging is
580 affected: the continued delivery log lines do not have any DNSSEC, TLS
581 Certificate or OCSP information. TLS cipher information is still logged.
583 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
584 identical IP addresses on different listening ports. Will also affect
587 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
588 add noisy ifdef guards to special-case this sillyness.
589 Patch from Bernd Kuhls.
591 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
592 than 255 are no longer allowed.
594 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
595 Disable the session-cache too, which might reduce our load. Since we
596 currrectly use a new context for every connection, both as server and
597 client, there is no benefit for these.
598 GnuTLS appears to not support tickets server-side by default (we don't
599 call gnutls_session_ticket_enable_server()) but client side is enabled
600 by default on recent versions (3.1.3 +) unless the PFS priority string
603 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
604 <https://reproducible-builds.org/specs/source-date-epoch/>.
606 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
607 the check for any unsuccessful recipients did not notice the limit, and
608 erroneously found still-pending ones.
610 JH/08 Pipeline CHUNKING command and data together, on kernels that support
611 MSG_MORE. Only in-clear (not on TLS connections).
613 JH/09 Avoid using a temporary file during transport using dkim. Unless a
614 transport-filter is involved we can buffer the headers in memory for
615 creating the signature, and read the spool data file once for the
616 signature and again for transmission.
618 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
619 4.77 as the kernel support then wasn't solid, having issues in 64bit
620 mode. Now, it's been long enough. Add support for FreeBSD also.
622 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
623 case where the routing stage had gathered several addresses to send to
624 a host before calling the transport for the first, we previously failed
625 to close down TLS in the old transport process before passing the TCP
626 connection to the new process. The new one sent a STARTTLS command
627 which naturally failed, giving a failed delivery and bloating the retry
628 database. Investigation and fix prototype from Wolfgang Breyha.
630 JH/12 Fix check on SMTP command input synchronisation. Previously there were
631 false-negatives in the check that the sender had not preempted a response
632 or prompt from Exim (running as a server), due to that code's lack of
633 awareness of the SMTP input buffering.
635 PP/04 Add commandline_checks_require_admin option.
636 Exim drops privileges sanely, various checks such as -be aren't a
637 security problem, as long as you trust local users with access to their
638 own account. When invoked by services which pass untrusted data to
639 Exim, this might be an issue. Set this option in main configuration
640 AND make fixes to the calling application, such as using `--` to stop
643 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
644 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
645 responses to those, into a single TLS record each way (this usually means
646 a single packet). As a side issue, smtp_enforce_sync now works on TLS
649 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
650 affects you only if you're dancing at the edge of the param size limits.
651 If you are, and this message makes sense to you, then: raise the
652 configured limit or use OpenSSL 1.1. Nothing we can do for older
655 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
656 cmdline element to get the documented default one. Previously it was
659 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
660 are unsuable for non-message-specific reasons - eg. connection timeout,
661 banner-time rejection.
663 JH/16 Fix logging of delivery remote port, when specified by router, under
666 PP/06 Repair manualroute's ability to take options in any order, even if one
667 is the name of a transport.
670 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
672 JH/17 Change the list-building routines interface to use the expanding-string
673 triplet model, for better allocation and copying behaviour.
675 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
676 Previously it was constructed the first time a possibly-matching string
677 was met in the configuration file input during startup; now it is done
680 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
681 compatible one, to avoid the (poorly documented) possibility of a config
682 file in the working directory redirecting the DB files, possibly correpting
683 some existing file. CVE-2017-10140 assigned for BDB.
685 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
686 cache-hot. Previously, although the result was properly cached, the
687 initial verify call returned a defer.
689 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
690 the main verify for receipient in uncached-mode.
692 JH/22 Retire historical build files to an "unsupported" subdir. These are
693 defined as "ones for which we have no current evidence of testing".
695 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
696 if present. Previously it was ignored.
698 JH/24 Start using specified-initialisers in C structure init coding. This is
699 a C99 feature (it's 2017, so now considered safe).
701 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
702 if was a fixed-sized field and bitmask ops via macros; it is now more
705 PP/07 GitHub PR 56: Apply MariaDB build fix.
706 Patch provided by Jaroslav Škarvada.
708 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
709 during Coverity cleanups [4.87 JH/47]
710 Diagnosis and fix provided by Michael Fischer v. Mollard.
712 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
713 the right size to place the terminating semicolon on its own folded
714 line, the header hash was calculated to an incorrect value thanks to
715 the (relaxed) space the fold became.
717 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
718 and confused the parent.
720 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
721 which could crash as a result. This could lead to undeliverable messages.
723 JH/28 Logging: "next input sent too soon" now shows where input was truncated
726 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
727 matters on fast-turnover and PID-randomising systems, which were getting
728 out-of-order delivery.
730 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
731 a possibly-overlapping copy. The symptom was that "Remote host closed
732 connection in response to HELO" was logged instead of the actual 4xx
735 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
736 Previously only that bufferd was discarded, resulting in SYMTP command
739 JH/32 DKIM: when a message has multiple signatures matching an identity given
740 in dkim_verify_signers, run the dkim acl once for each. Previously only
741 one run was done. Bug 2189.
743 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
744 "panic the current process" to "deliberately defer". The panic log is
745 still written with the problem list name; the mail and reject logs now
746 get a temp-reject line for the message that was being handled, saying
747 something like "domains check lookup or other defer". The SMTP 451
748 message is still "Temporary local problem".
750 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
751 A crafted sequence of BDAT commands could result in in-use memory beeing
752 freed. CVE-2017-16943.
754 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
755 from SMTP input. Previously it was always done; now only done for DATA
756 and not BDAT commands. CVE-2017-16944.
758 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
759 to the message (such as an overlong header line). Previously this was
760 not done and we did not exit BDAT mode. Followon from the previous item
761 though a different problem.
767 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
768 than -2003 did; needs libidn2 in addition to libidn.
770 JH/02 The path option on a pipe transport is now expanded before use.
772 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
773 Patch provided by "Björn", documentation fix added too.
775 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
776 missing a wire-to-host endian conversion.
778 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
779 close after a BDAT command line could be taken as a following command,
780 giving a synch failure. Fix by only checking for synch immediately
781 before acknowledging the chunk.
783 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
784 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
785 macro. Patches provided by Josh Soref.
787 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
788 Previously we did not; the RFC seems ambiguous and VRFY is not listed
789 by IANA as a service extension. However, John Klensin suggests that we
792 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
793 the dkim code may be unix-mode line endings rather than smtp wire-format
794 CRLF, so prepend a CR to any bare LF.
796 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
797 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
799 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
800 the first were themselves being wrongly included in the feed into dkim
801 processing; with most chunk sizes in use this resulted in an incorrect
802 body hash calculated value.
804 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
805 DKIM signature block, for verification. Although advised against by
806 standards it is specifically not ruled illegal.
808 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
810 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
811 missing a body hash (the bh= tag).
813 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
814 It seems that HAProxy sends the Proxy Protocol information in clear and
815 only then does a TLS startup, so do the same.
817 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
818 TCP connections (such as for Spamd) unless the daemon successfully set
819 Fast Open mode on its listening sockets. This fixes breakage seen on
820 too-old kernels or those not configured for Fast Open, at the cost of
821 requiring both directions being enabled for TFO, and TFO never being used
822 by non-daemon-related Exim processes.
824 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
825 endings, at least on the first header line. Try to canonify any that get
826 past that check, despite the cost.
828 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
829 now limited to an arbitrary five deep, while parsing addresses with the
830 strip_excess_angle_brackets option enabled.
832 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
833 instead leave the unprompted TLS handshake in socket buffer for the
834 TLS library to consume.
836 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
838 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
840 JH/16 Drop variables when they go out of scope. Memory management drops a whole
841 region in one operation, for speed, and this leaves assigned pointers
842 dangling. Add checks run only under the testsuite which checks all
843 variables at a store-reset and panics on a dangling pointer; add code
844 explicitly nulling out all the variables discovered. Fixes one known
845 bug: a transport crash, where a dangling pointer for $sending_ip_address
846 originally assigned in a verify callout, is re-used.
848 PP/06 Drop '.' from @INC in various Perl scripts.
850 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
852 PP/08 Reduce a number of compilation warnings under clang; building with
853 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
854 should be warning-free.
856 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
858 HS/01 Fix portability problems introduced by PP/08 for platforms where
859 realloc(NULL) is not equivalent to malloc() [SunOS et al].
861 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
862 chunk. This allows us to accept broken chunked messages. We need a more
863 general solution here.
865 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
866 already-broken messages in the queue.
868 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
870 JH/19 Fix reference counting bug in routing-generated-address tracking.
876 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
877 supports it and a size is available (ie. the sending peer gave us one).
879 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
880 years of being deprecated). The replacements are the ACLs
881 acl_smtp_mime and acl_not_smtp_mime.
883 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
884 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
885 or even an in-clear connection were permitted. Now, if the host lookup
886 was dnssec and dane was requested then the host is only used if the
887 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
888 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
889 if one fails this test.
890 This means that a poorly-configured remote DNS will make it incommunicado;
891 but it protects against a DNS-interception attack on it.
893 JH/04 Bug 1810: make continued-use of an open smtp transport connection
894 non-noisy when a race steals the message being considered.
896 JH/05 If main configuration option tls_certificate is unset, generate a
897 self-signed certificate for inbound TLS connections.
899 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
900 in rewrites and routers.
902 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
903 and logged a warning sing 4.83; now they are a configuration file error.
905 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
906 (lacking @domain). Apply the same qualification processing as RCPT.
908 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
910 JH/10 Support ${sha256:} applied to a string (as well as the previous
913 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
914 a cutthrough deliver is pending, as we always want to make a connection.
915 This also avoids re-routing the message when later placing the cutthrough
916 connection after a verify cache hit.
917 Do not update it with the verify result either.
919 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
920 when routing results in more than one destination address.
922 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
923 signing (which inhibits the cutthrough capability). Previously only
924 the presence of an option was tested; now an expansion evaluating as
925 empty is permissible (obviously it should depend only on data available
926 when the cutthrough connection is made).
928 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
929 the relevant preceding SMTP command did not note the pipelining mode.
931 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
932 Previously they were not counted.
934 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
935 as one having no matching records. Previously we deferred the message
936 that needed the lookup.
938 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
939 distinguished as "(=".
941 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
942 for missing MX records. Previously it only worked for missing A records.
944 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
946 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
947 after the data-go-ahead and data-ack. Patch from Jason Betts.
949 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
950 even for a "none" policy. Patch from Tony Meyer.
952 JH/22 Fix continued use of a connection for further deliveries. If a port was
953 specified by a router, it must also match for the delivery to be
956 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
957 When one of the recipients of a message was unsuitable for the connection
958 (has no matching addresses), we lost track of needing to mark it
959 deferred. As a result mail would be lost.
961 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
963 JH/25 Decoding ACL controls is now done using a binary search; the source code
964 takes up less space and should be simpler to maintain. Merge the ACL
965 condition decode tables also, with similar effect.
967 JH/26 Fix problem with one_time used on a redirect router which returned the
968 parent address unchanged. A retry would see the parent address marked as
969 delivered, so not attempt the (identical) child. As a result mail would
972 JH/27 Fix a possible security hole, wherein a process operating with the Exim
973 UID can gain a root shell. Credit to http://www.halfdog.net/ for
974 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
977 JH/28 Enable {spool,log} filesystem space and inode checks as default.
978 Main config options check_{log,spool}_{inodes,space} are now
979 100 inodes, 10MB unless set otherwise in the configuration.
981 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
982 Previously it only applied to the main-section connection policy
985 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
987 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
988 by me. Added RFC7919 DH primes as an alternative.
990 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
991 are not in the system include path.
993 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
994 GnuTLS, when a session startup failed (eg because the client disconnected)
995 Exim did stdio operations after fclose. This was exposed by a recent
996 change which nulled out the file handle after the fclose.
998 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
999 signed directly by the cert-signing cert, rather than an intermediate
1000 OCSP-signing cert. This is the model used by LetsEncrypt.
1002 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1004 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1005 an incoming connection.
1007 HS/02 Bug 1802: Do not half-close the connection after sending a request
1010 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1011 fallback to "prime256v1".
1013 JH/34 SECURITY: Use proper copy of DATA command in error message.
1014 Could leak key material. Remotely exploitable. CVE-2016-9963.
1020 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1021 and 3.4.4 - once the server is enabled to respond to an OCSP request
1022 it does even when not requested, resulting in a stapling non-aware
1023 client dropping the TLS connection.
1025 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1026 support variable-length bit vectors. No functional change.
1028 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1029 The I= interface field on outgoing lines is now after the H= remote
1030 host field, same as incoming lines. There is a separate
1031 outgoing_interface log selector which allows you to disable the
1034 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1035 If not running log_selector +smtp_connection the mainlog would be held
1036 open indefinitely after a "too many connections" event, including to a
1037 deleted file after a log rotate. Leave the per net connection logging
1038 leaving it open for efficiency as that will be quickly detected by the
1039 check on the next write.
1041 HS/01 Bug 1671: Fix post transport crash.
1042 Processing the wait-<transport> messages could crash the delivery
1043 process if the message IDs didn't exist for some reason. When
1044 using 'split_spool_directory=yes' the construction of the spool
1045 file name failed already, exposing the same netto behaviour.
1047 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1048 mime_regex ACL conditions.
1050 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1051 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1052 to HELO, local diagnostic string.
1054 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1055 log line to debug. Even when configured with a tls authenticator many
1056 client connections are expected to not authenticate in this way, so
1057 an authenticate fail is not an error.
1059 HS/02 Add the Exim version string to the process info. This way exiwhat
1060 gives some more detail about the running daemon.
1062 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1063 matter for fast-change records such as DNSBLs.
1065 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1066 retry record, even if constant. There may be multiple transports with
1067 different interface settings and the retry behaviour needs to be kept
1070 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1072 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1074 JH/10 Bug 840: fix log_defer_output option of pipe transport
1076 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1077 hosts_randomize. This matters a lot when combined with mua_wrapper.
1079 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1080 ${quote_pgsql:<string>} operator.
1082 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1084 JH/14 Change header file name for freeradius-client. Relevant if compiling
1085 with Radius support; from the Gentoo tree and checked under Fedora.
1087 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1089 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1090 option transport_filter, meaning no filtering.
1092 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1094 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1095 defaults to "*" (all hosts). The variable is now available when not built
1096 with TLS, default unset, mainly to enable keeping the testsuite sane.
1097 If a server certificate is not supplied (via tls_certificate) an error is
1098 logged, and clients will find TLS connections fail on startup. Presumably
1099 they will retry in-clear.
1100 Packagers of Exim are strongly encouraged to create a server certificate
1101 at installation time.
1103 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1104 with the $config_file variable.
1106 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1107 in transport context, after the attempt, and per-recipient. The latter type
1108 is per host attempted. The event data is the error message, and the errno
1109 information encodes the lookup type (A vs. MX) used for the (first) host,
1110 and the trailing two digits of the smtp 4xx response.
1112 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1113 to write to mainlog (or rejectlog, paniclog) in the window between file
1114 creation and permissions/ownership being changed. Particularly affects
1115 installations where exicyclog is run as root, rather than exim user;
1116 result is that the running daemon panics and dies.
1118 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1120 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1121 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1122 "pri" and "weight". Note that the previous implicit priority given by the
1123 list order is no longer honoured.
1125 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1126 for DKIM processing.
1128 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1129 by defining SUPPORT_SOCKS.
1131 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1132 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1133 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1134 variables are renamed to proxy_{local,external}_{address,port}.
1136 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1137 for a build by defining SUPPORT_I18N
1139 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1140 of the query string, and make ${quote_redis:} do that quoting.
1142 JH/29 Move Events support from Experimental to mainline, enabled by default
1143 and removable for a build by defining DISABLE_EVENT.
1145 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1147 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1148 cached by the daemon.
1150 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1151 by defining LOOKUP_REDIS. The libhiredis library is required.
1153 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1154 keys are given for lookup.
1156 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1157 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1158 only supported when built with TLS support. The PolarSSL SHA routines
1159 are still used when the TLS library is too old for convenient support.
1161 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1162 openssl_options), for security. OpenSSL forces this from version 1.1.0
1163 server-side so match that on older versions.
1165 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1166 allocation for $value could be released as the expansion processing
1167 concluded, but leaving the global pointer active for it.
1169 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1170 and to use the domains and local_parts ACL conditions.
1172 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1173 incorrectly not doubled on cutthrough transmission, hence seen as a
1174 body-termination at the receiving system - resulting in truncated mails.
1175 Commonly the sender saw a TCP-level error, and retransmitted the message
1176 via the normal store-and-forward channel. This could result in duplicates
1177 received - but deduplicating mailstores were liable to retain only the
1178 initial truncated version.
1180 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1182 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1184 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1185 we're in there, support oversigning also; bug 1309.
1187 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1189 HS/04 Add support for keep_environment and add_environment options.
1191 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1192 either intentional arithmetic overflow during PRNG, or testing config-
1195 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1196 delivery resulted in actual delivery. Cancel cutthrough before DATA
1199 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1200 rejects a recipient: pass the reject to the originator.
1202 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1203 Many were false-positives and ignorable, but it's worth fixing the
1206 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1207 for the new environment-manipulation done at startup. Move the routines
1208 from being local to tls.c to being global via the os.c file.
1210 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1211 an extract embedded as result-arg for a map, the first arg for extract
1212 is unavailable so we cannot tell if this is a numbered or keyed
1213 extraction. Accept either.
1219 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1222 JH/02 The smtp transport option "multi_domain" is now expanded.
1224 JH/03 The smtp transport now requests PRDR by default, if the server offers
1227 JH/04 Certificate name checking on server certificates, when exim is a client,
1228 is now done by default. The transport option tls_verify_cert_hostnames
1229 can be used to disable this per-host. The build option
1230 EXPERIMENTAL_CERTNAMES is withdrawn.
1232 JH/05 The value of the tls_verify_certificates smtp transport and main options
1233 default to the word "system" to access the system default CA bundle.
1234 For GnuTLS, only version 3.0.20 or later.
1236 JH/06 Verification of the server certificate for a TLS connection is now tried
1237 (but not required) by default. The verification status is now logged by
1238 default, for both outbound TLS and client-certificate supplying inbound
1241 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1244 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1245 Status Notification (bounce) messages are now MIME format per RFC 3464.
1246 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1247 under the control of the dsn_advertise_hosts option, and routers may
1248 have a dsn_lasthop option.
1250 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1251 default, modifiable by a malware= option. The list separator for
1252 the options can now be changed in the usual way. Bug 68.
1254 JH/10 The smtp_receive_timeout main option is now expanded before use.
1256 JH/11 The incoming_interface log option now also enables logging of the
1257 local interface on delivery outgoing connections.
1259 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1260 if the interface and destination host and port all match.
1262 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1265 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1266 Patch from Andrew Lewis.
1268 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1269 now supports optional time-restrictions, weighting, and priority
1270 modifiers per server. Patch originally by <rommer@active.by>.
1272 JH/16 The spamd_address main option now supports a mixed list of local
1273 and remote servers. Remote servers can be IPv6 addresses, and
1274 specify a port-range.
1276 JH/17 Bug 68: The spamd_address main option now supports an optional
1277 timeout value per server.
1279 JH/18 Bug 1581: Router and transport options headers_add/remove can
1280 now have the list separator specified.
1282 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1285 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1288 JH/21 Support for the A6 type of dns record is withdrawn.
1290 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1291 rather than the verbs used.
1293 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1294 from 255 to 1024 chars.
1296 JH/24 Verification callouts now attempt to use TLS by default.
1298 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1299 are generic router options now. The defaults didn't change.
1301 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1302 Original patch from Alexander Shikoff, worked over by JH.
1304 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1305 files if ZCAT_COMMAND is not executable.
1307 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1309 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1311 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1312 Normally benign, it bites when the pair was led to by a CNAME;
1313 modern usage is to not canonicalize the domain to a CNAME target
1314 (and we were inconsistent anyway for A-only vs AAAA+A).
1316 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1318 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1319 when evaluating $sender_host_dnssec.
1321 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1322 $sender_helo_dnssec variable.
1324 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1326 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1328 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1330 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1331 documented as working, but never had. Support all but $spam_report.
1333 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1334 added for tls authenticator.
1336 HS/03 Add perl_taintmode main config option
1342 TL/01 When running the test suite, the README says that variables such as
1343 no_msglog_check are global and can be placed anywhere in a specific
1344 test's script, however it was observed that placement needed to be near
1345 the beginning for it to behave that way. Changed the runtest perl
1346 script to read through the entire script once to detect and set these
1347 variables, reset to the beginning of the script, and then run through
1348 the script parsing/test process like normal.
1350 TL/02 The BSD's have an arc4random API. One of the functions to induce
1351 adding randomness was arc4random_stir(), but it has been removed in
1352 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1353 function when detected.
1355 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1356 cause callback expansion.
1358 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1359 syntax errors in an expansion can be treated as a string instead of
1360 logging or causing an error, due to the internal use of bool_lax
1361 instead of bool when processing it.
1363 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1364 server certificates when making smtp deliveries.
1366 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1368 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1370 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1372 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1373 Merged patch from Sebastian Wiedenroth.
1375 JH/05 Fix results-pipe from transport process. Several recipients, combined
1376 with certificate use, exposed issues where response data items split
1377 over buffer boundaries were not parsed properly. This eventually
1378 resulted in duplicates being sent. This issue only became common enough
1379 to notice due to the introduction of connection certificate information,
1380 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1382 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1383 size buffer was used, resulting in syntax errors when an expansion
1386 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1387 version 3.3.6 or later.
1389 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1390 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1391 both become "event_action", the variables become $event_name, $event_data
1392 and $event_defer_errno. There is a new variable $verify_mode, usable in
1393 routers, transports and related events. The tls:cert event is now also
1394 raised for inbound connections, if the main configuration event_action
1397 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1398 early OCSP support, but no stapling (appears to be less than 1.0.0).
1400 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1401 server certificate names available under the smtp transport option
1402 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1405 JH/10 Time-related extraction expansions from certificates now use the main
1406 option "timezone" setting for output formatting, and are consistent
1407 between OpenSSL and GnuTLS compilations. Bug 1541.
1409 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1410 encoded parameter in the incoming message. Bug 1558.
1412 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1413 include certificate info, eximon was claiming there were spoolfile
1416 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1418 JH/14 Log delivery-related information more consistently, using the sequence
1419 "H=<name> [<ip>]" wherever possible.
1421 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1422 are problematic for Debian distribution, omit them from the release
1425 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1427 JH/16 Fix string representation of time values on 64bit time_t architectures.
1430 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1431 output list separator was used.
1436 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1437 checkers that were complaining about end of non-void function with no
1440 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1441 This was a regression introduced in 4.83 by another bugfix.
1443 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1445 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1446 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1452 TF/01 Correctly close the server side of TLS when forking for delivery.
1454 When a message was received over SMTP with TLS, Exim failed to clear up
1455 the incoming connection properly after forking off the child process to
1456 deliver the message. In some situations the subsequent outgoing
1457 delivery connection happened to have the same fd number as the incoming
1458 connection previously had. Exim would try to use TLS and fail, logging
1459 a "Bad file descriptor" error.
1461 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1462 utilities have not been installed.
1464 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1465 temporary space as the ACL may create new global variables.
1467 TL/01 LDAP support uses per connection or global context settings, depending
1468 upon the detected version of the libraries at build time.
1470 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1471 to extract and use the src ip:port in logging and expansions as if it
1472 were a direct connection from the outside internet. PPv2 support was
1473 updated based on HAProxy spec change in May 2014.
1475 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1477 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1478 Properly escape header and check for NULL return.
1480 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1483 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1485 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1486 characters in header names, implemented as a verify condition.
1487 Contributed by Michael Fischer v. Mollard.
1489 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1490 results permerror and temperror. Previous values are deprecated but
1491 still accepted. In a future release, err_perm and err_temp will be
1492 completely removed, which will be a backward incompatibility if the
1493 ACL tests for either of these two old results. Patch contributed by
1494 user bes-internal on the mailing list.
1496 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1498 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1499 selectors, in both main and reject logs.
1501 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1504 JH/07 Add malware type "sock" for talking to simple daemon.
1506 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1508 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1509 routers/transports under cutthrough routing.
1511 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1512 numbers. Touch up "bool" conditional to keep the same definition.
1514 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1516 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1518 JH/12 Expand items in router/transport headers_add or headers_remove lists
1519 individually rather than the list as a whole. Bug 1452.
1521 Required for reasonable handling of multiple headers_ options when
1522 they may be empty; requires that headers_remove items with embedded
1523 colons must have them doubled (or the list-separator changed).
1525 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1526 view the policy declared in the DMARC record. Currently, $dmarc_status
1527 is a combined value of both the record presence and the result of the
1530 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1532 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1533 dnslookup router and the smtp transport (applying to the forward
1536 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1537 of ldap servers used for a specific lookup. Patch provided by Heiko
1540 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1541 New variable $lookup_dnssec_authenticated for observability.
1543 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1544 Patch submitted by Lars Timman.
1546 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1548 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1549 Requires trusted mode and valid format message id, aborts otherwise.
1550 Patch contributed by Heiko Schlichting.
1552 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1553 certextract with support for various fields. Bug 1358.
1555 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1556 is requested by default, modifiable by smtp transport option
1559 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1560 operate on certificate variables to give certificate fingerprints
1561 Also new ${sha256:cert_variable}.
1563 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1565 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1566 Christian Aistleitner.
1568 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1570 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1571 file. Patch from Wolfgang Breyha.
1573 JH/25 Expand the coverage of the delivery $host and $host_address to
1574 client authenticators run in verify callout. Bug 1476.
1576 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1577 align with daemon_smtp_ports. Bug 72.
1579 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1580 support and error reporting did not work properly.
1582 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1583 and is readable. Patch from Andrew Colin Kissa.
1585 TL/14 Enhance documentation of ${run expansion and how it parses the
1586 commandline after expansion, particularly in the case when an
1587 unquoted variable expansion results in an empty value.
1589 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1591 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1592 by renaming away from it. Addresses GH issue 15, reported by
1595 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1597 TL/15 SECURITY: prevent double expansion in math comparison functions
1598 (can expand unsanitized data). Not remotely exploitable.
1605 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1607 PP/02 Make -n do something, by making it not do something.
1608 When combined with -bP, the name of an option is not output.
1610 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1613 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1614 $sender_host_name and config options to manage this, and basic check
1617 PP/05 DSCP support for outbound connections and control modifier for inbound.
1619 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1620 (Only plugin which currently uses this is kerberos4, which nobody should
1621 be using, but we should make it available and other future plugins might
1622 conceivably use it, even though it would break NAT; stuff *should* be
1623 using channel bindings instead).
1625 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1626 name; added for Sendmail compatibility; requires admin caller.
1627 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1628 just ignore it); requires trusted caller.
1629 Also parse but ignore: -Ac -Am -X<logfile>
1632 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1634 TL/02 Add +smtp_confirmation as a default logging option.
1636 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1637 Patch by Magnus Holmgren from 2007-02-20.
1639 TL/04 Bugzilla 1281 - Spec typo.
1640 Bugzilla 1283 - Spec typo.
1641 Bugzilla 1290 - Spec grammar fixes.
1643 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1645 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1647 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1648 mailing list by by Dmitry Isaikin.
1650 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1652 JH/02 Support "G" suffix to numbers in ${if comparisons.
1654 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1656 NM/01 Bugzilla 1197 - Spec typo
1657 Bugzilla 1196 - Spec examples corrections
1659 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1661 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1662 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1665 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1666 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1668 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1669 "acl {{name}{arg}...}", and optional args on acl condition
1672 JH/05 Permit multiple router/transport headers_add/remove lines.
1674 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1676 JH/07 Avoid using a waiting database for a single-message-only transport.
1677 Performance patch from Paul Fisher. Bugzilla 1262.
1679 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1682 JH/09 Add $headers_added variable, with content from use of ACL modifier
1683 add_header (but not yet added to the message). Bugzilla 199.
1685 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1686 Pulled from Bugzilla 817 by Wolfgang Breyha.
1688 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1690 (nb: this is the same fix as in Exim 4.80.1)
1692 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1695 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1698 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1700 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1701 advertise SMTP AUTH mechanism to us, instead of a generic
1702 protocol violation error. Also, make Exim more robust to bad
1703 data from the Dovecot auth socket.
1705 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1707 When a queue runner is handling a message, Exim first routes the
1708 recipient addresses, during which it prunes them based on the retry
1709 hints database. After that it attempts to deliver the message to
1710 any remaining recipients. It then updates the hints database using
1713 So if a recipient address works intermittently, it can get repeatedly
1714 deferred at routing time. The retry hints record remains fresh so the
1715 address never reaches the final cutoff time.
1717 This is a fairly common occurrence when a user is bumping up against
1718 their storage quota. Exim had some logic in its local delivery code
1719 to deal with this. However it did not apply to per-recipient defers
1720 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1722 This change adds a proper retry rule check during routing so that the
1723 final cutoff time is checked against the message's age. We only do
1724 this check if there is an address retry record and there is not a
1725 domain retry record; this implies that previous attempts to handle
1726 the address had the retry_use_local_parts option turned on. We use
1727 this as an approximation for the destination being like a local
1728 delivery, as in LMTP.
1730 I suspect this new check makes the old local delivery cutoff check
1731 redundant, but I have not verified this so I left the code in place.
1733 TF/02 Correct gecos expansion when From: is a prefix of the username.
1735 Test 0254 submits a message to Exim with the header
1739 When I ran the test suite under the user fanf2, Exim expanded
1740 the header to contain my full name, whereas it should have added
1741 a Resent-Sender: header. It erroneously treats any prefix of the
1742 username as equal to the username.
1744 This change corrects that bug.
1746 GF/01 DCC debug and logging tidyup
1747 Error conditions log to paniclog rather than rejectlog.
1748 Debug lines prefixed by "DCC: " to remove any ambiguity.
1750 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1752 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1753 Bug spotted by Jeremy Harris; was flawed since initial commit.
1754 Would have resulted in OCSP responses post-SNI triggering an Exim
1755 NULL dereference and crash.
1757 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1759 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1760 Bug detection, analysis and fix by Samuel Thibault.
1761 Bugzilla 1331, Debian bug #698092.
1763 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1765 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1766 Server implementation by Todd Lyons, client by JH.
1767 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1768 config variable "prdr_enable" controls whether the server
1769 advertises the facility. If the client requests PRDR a new
1770 acl_data_smtp_prdr ACL is called once for each recipient, after
1771 the body content is received and before the acl_smtp_data ACL.
1772 The client is controlled by both of: a hosts_try_prdr option
1773 on the smtp transport, and the server advertisement.
1774 Default client logging of deliveries and rejections involving
1775 PRDR are flagged with the string "PRDR".
1777 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1778 fclose(). Diagnosis by Todd Lyons.
1780 PP/17 Update configure.default to handle IPv6 localhost better.
1781 Patch by Alain Williams (plus minor tweaks).
1784 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1785 This is now consistent with GnuTLS, and is now documented: the
1786 previous undocumented portable approach to treating the option as
1787 unset was to force an expansion failure. That still works, and
1788 an empty string is now equivalent.
1790 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1791 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1792 not performing validation itself.
1794 PP/20 Added force_command boolean option to pipe transport.
1795 Patch from Nick Koston, of cPanel Inc.
1797 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1800 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1802 PP/21 Fix eximon continuous updating with timestamped log-files.
1803 Broken in a format-string cleanup in 4.80, missed when I repaired the
1804 other false fix of the same issue.
1805 Report and fix from Heiko Schlichting.
1808 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1809 Report from Prashanth Katuri.
1811 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1812 It's SecureTransport, so affects any MacOS clients which use the
1813 system-integrated TLS libraries, including email clients.
1815 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1816 using a MIME ACL for non-SMTP local injection.
1817 Report and assistance in diagnosis by Warren Baker.
1819 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1821 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1823 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1824 last id that failed so it may be referenced in subsequent ACL's.
1826 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1829 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1830 ldap library initialization, allowing self-signed CA's to be
1831 used. Also properly sets require_cert option later in code by
1832 using NULL (global ldap config) instead of ldap handle (per
1833 session). Bug diagnosis and testing by alxgomz.
1835 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1836 the src/util/ subdirectory.
1838 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1839 renamed to Transport Post Delivery Action by Jeremy Harris, as
1842 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1843 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1844 redis_servers = needs to be configured which will be used by the redis
1845 lookup. Patch from Warren Baker, of The Packet Hub.
1847 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1849 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1850 hostname or reverse DNS when processing a host list. Used suggestions
1851 from multiple comments on this bug.
1853 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1855 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1856 Missed a few lines, added it to make the runtest require no keyboard
1859 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1860 contains upper case chars. Make router use caseful_local_part.
1862 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1863 support when GnuTLS has been built with p11-kit.
1869 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1871 This, or similar/improved, will also be change PP/11 of 4.82.
1877 PP/01 Handle short writes when writing local log-files.
1878 In practice, only affects FreeBSD (8 onwards).
1879 Bugzilla 1053, with thanks to Dmitry Isaikin.
1881 NM/01 Bugzilla 949 - Documentation tweak
1883 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1886 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1888 PP/02 Implemented gsasl authenticator.
1890 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1892 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1893 `pkg-config foo` for cflags/libs.
1895 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1896 with rest of GSASL and with heimdal_gssapi.
1898 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1899 `pkg-config foo` for cflags/libs for the TLS implementation.
1901 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1902 properties get this fed in as external SSF. A number of robustness
1903 and debugging improvements to the cyrus_sasl authenticator.
1905 PP/08 cyrus_sasl server now expands the server_realm option.
1907 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1908 Patch by Jeremy Harris.
1910 PP/10 Added dbmjz lookup type.
1912 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1914 PP/12 MAIL args handles TAB as well as SP, for better interop with
1915 non-compliant senders.
1916 Analysis and variant patch by Todd Lyons.
1918 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1919 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1920 Patch from Dirk Mueller <dmueller@suse.com>
1922 PP/13 tls_peerdn now print-escaped for spool files.
1923 Observed some $tls_peerdn in wild which contained \n, which resulted
1924 in spool file corruption.
1926 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1927 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1928 or write after TLS renegotiation, which otherwise led to messages
1931 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1932 as a tracking header (ie: a signed header comes before the signature).
1933 Patch from Wolfgang Breyha.
1935 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1936 comma-sep list; embedded commas doubled.
1938 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1940 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1942 Report and patch from Dmitry Banschikov.
1944 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1945 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1946 failures. We appear to now support TLS1.1+ with Exim.
1948 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1949 lets Exim select keys and certificates based upon TLS SNI from client.
1950 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1951 before an outbound SMTP session. New log_selector, +tls_sni.
1953 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1954 NULL dereference. Report and patch from Alun Jones.
1956 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1957 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1958 Not seeing resolver debug output on NetBSD, but suspect this is a
1959 resolver implementation change.
1961 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1962 Left warnings. Added "eximon gdb" invocation mode.
1964 PP/21 Defaulting "accept_8bitmime" to true, not false.
1966 PP/22 Added -bw for inetd wait mode support.
1968 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1969 locate the relevant includes and libraries. Made this the default.
1971 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1972 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1974 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1975 This may cause build issues on older platforms.
1977 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1978 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1979 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1980 Added SNI support via GnuTLS too.
1981 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1983 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1985 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1986 Applied second patch from Janne, implementing suggestion to default
1987 multiple-strings-in-record handling to match SPF spec.
1989 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1991 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1992 read-only, out of scope).
1993 Patch from Wolfgang Breyha, report from Stuart Northfield.
1995 PP/29 Fix three issues highlighted by clang analyser static analysis.
1996 Only crash-plausible issue would require the Cambridge-specific
1997 iplookup router and a misconfiguration.
1998 Report from Marcin Mirosław.
2000 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2002 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2003 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2004 As part of this, removing so much warning spew let me fix some minor
2005 real issues in debug logging.
2007 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2008 assignment on my part. Fixed.
2010 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2011 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2012 Janne Snabb (who went above and beyond: thank you).
2014 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2015 string otherwise requires a connection and a bunch more work and it's
2016 relatively easy to get wrong. Should also expose TLS library linkage
2019 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2020 64-bit ${eval} (JH/03).
2022 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2023 GNU libc to support some of the 64-bit stuff, should not lead to
2024 conflicts. Defined before os.h is pulled in, so if a given platform
2025 needs to override this, it can.
2027 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2028 protection layer was required, which is not implemented.
2029 Bugzilla 1254, patch from Wolfgang Breyha.
2031 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2032 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2033 tls_dhparam take prime identifiers. Also unbreak combination of
2034 OpenSSL+DH_params+TLSSNI.
2036 PP/39 Disable SSLv2 by default in OpenSSL support.
2042 PP/01 Solaris build fix for Oracle's LDAP libraries.
2043 Bugzilla 1109, patch from Stephen Usher.
2045 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2047 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2050 TF/02 Fix a couple more cases where we did not log the error message
2051 when unlink() failed. See also change 4.74-TF/03.
2053 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2054 lock up or crash if it happened to be inside a call to libc when it
2055 got a SIGUSR1 from exiwhat.
2057 The SIGUSR1 handler appends the current process status to the process
2058 log which is later printed by exiwhat. It used to use the general
2059 purpose logging code to do this, but several functions it calls are
2060 not safe for signals.
2062 The new output code in the SIGUSR1 handler is specific to the process
2063 log, and simple enough that it's easy to inspect for signal safety.
2064 Removing some special cases also simplifies the general logging code.
2065 Removing the spurious timestamps from the process log simplifies
2068 TF/04 Improved ratelimit ACL condition.
2070 The /noupdate option has been deprecated in favour of /readonly which
2071 has clearer semantics. The /leaky, /strict, and /readonly update modes
2072 are mutually exclusive. The update mode is no longer included in the
2073 database key; it just determines when the database is updated. (This
2074 means that when you upgrade Exim will forget old rate measurements.)
2076 Exim now checks that the per_* options are used with an update mode that
2077 makes sense for the current ACL. For example, when Exim is processing a
2078 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2079 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2080 must specify per_mail/readonly. If you omit the update mode it defaults to
2081 /leaky where that makes sense (as before) or /readonly where required.
2083 The /noupdate option is now undocumented but still supported for
2084 backwards compatibility. It is equivalent to /readonly except that in
2085 ACLs where /readonly is required you may specify /leaky/noupdate or
2086 /strict/noupdate which are treated the same as /readonly.
2088 A useful new feature is the /count= option. This is a generalization
2089 of the per_byte option, so that you can measure the throughput of other
2090 aggregate values. For example, the per_byte option is now equivalent
2091 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2093 The per_rcpt option has been generalized using the /count= mechanism
2094 (though it's more complicated than the per_byte equivalence). When it is
2095 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2096 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2097 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2098 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2099 Note that using per_rcpt with a non-readonly update mode in more than
2100 one ACL will cause the recipients to be double-counted. (The per_mail
2101 and per_byte options don't have this problem.)
2103 The handling of very low rates has changed slightly. If the computed rate
2104 is less than the event's count (usually one) then this event is the first
2105 after a long gap. In this case the rate is set to the same as this event's
2106 count, so that the first message of a spam run is counted properly.
2108 The major new feature is a mechanism for counting the rate of unique
2109 events. The new per_addr option counts the number of different
2110 recipients that someone has sent messages to in the last time period. It
2111 behaves like per_rcpt if all the recipient addresses are different, but
2112 duplicate recipient addresses do not increase the measured rate. Like
2113 the /count= option this is a general mechanism, so the per_addr option
2114 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2115 example, measure the rate that a client uses different sender addresses
2116 with the options per_mail/unique=$sender_address. There are further
2117 details in the main documentation.
2119 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2121 TF/06 Removed a few PCRE remnants.
2123 TF/07 Automatically extract Exim's version number from tags in the git
2124 repository when doing development or release builds.
2126 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2127 Bugzilla 879. Patch from Paul Fisher.
2129 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2130 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2133 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2135 PP/05 Variable $av_failed, true if the AV scanner deferred.
2136 Bugzilla 1078. Patch from John Horne.
2138 PP/06 Stop make process more reliably on build failure.
2139 Bugzilla 1087. Patch from Heiko Schlittermann.
2141 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2142 Bugzilla 1089. Patch from Heiko Schlittermann.
2144 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2145 Bugzilla 1131. Patch from Holger Weiß.
2147 PP/09 Handle IPv6 addresses with SPF.
2148 Bugzilla 860. Patch from Wolfgang Breyha.
2150 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2152 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2155 PP/11 match_* no longer expand right-hand-side by default.
2156 New compile-time build option, EXPAND_LISTMATCH_RHS.
2157 New expansion conditions, "inlist", "inlisti".
2159 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2161 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2163 PP/14 fix log_write() format string regression from TF/03.
2164 Bugzilla 1152. Patch from Dmitry Isaikin.
2170 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2172 PP/02 Harmonised TLS library version reporting; only show if debugging.
2173 Layout now matches that introduced for other libraries in 4.74 PP/03.
2175 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2177 PP/04 New "dns_use_edns0" global option.
2179 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2182 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2183 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2185 TK/01 Updated PolarSSL code to 0.14.2.
2186 Bugzilla 1097. Patch from Andreas Metzler.
2188 PP/07 Catch divide-by-zero in ${eval:...}.
2189 Fixes bugzilla 1102.
2191 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2194 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2195 format-string attack -- SECURITY: remote arbitrary code execution.
2197 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2198 time unintentionally subject to list matching rules, letting the header
2199 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2200 arbitrary string expansion). This allowed for information disclosure.
2202 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2203 INT_MIN/-1 -- value coerced to INT_MAX.
2209 NM/01 Workaround for PCRE version dependency in version reporting
2212 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2213 This fixes portability to compilers other than gcc, notably
2214 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2216 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2217 makefiles for portability to HP-UX and POSIX correctness.
2219 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2220 Also via indented variable definition in the Makefile.
2221 (Debugging by Oliver Heesakkers).
2223 PP/02 Restore caching of spamd results with expanded spamd_address.
2224 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2226 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2227 Improves build reliability. Fix from: Frank Elsner
2229 NM/02 Fix wide character breakage in the rfc2047 coding
2230 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2232 NM/03 Allow underscore in dnslist lookups
2233 Fixes bug 1026. Patch from Graeme Fowler
2235 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2236 Code patches from Adam Ciarcinski of NetBSD.
2238 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2241 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2242 is logged, to avoid truncation. Patch from John Horne.
2244 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2245 Patch from Jakob Hirsch.
2247 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2248 SQL string expansion failure details.
2249 Patch from Andrey Oktyabrski.
2251 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2252 Patch from Simon Arlott.
2254 PP/09 New lookups functionality failed to compile on old gcc which rejects
2255 extern declarations in function scope.
2256 Patch from Oliver Fleischmann
2258 PP/10 Use sig_atomic_t for flags set from signal handlers.
2259 Check getgroups() return and improve debugging.
2260 Fixed developed for diagnosis in bug 927 (which turned out to be
2263 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2264 Patch from Mark Zealey.
2266 PP/12 Bugzilla 1056: Improved spamd server selection.
2267 Patch from Mark Zealey.
2269 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2270 Based on patch from Heiko Schlittermann.
2272 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2273 Patch from Uwe Doering, sign-off by Michael Haardt.
2275 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2276 variable declaration deep within a block. Bug and patch from
2279 PP/15 lookups-Makefile IRIX compatibility coercion.
2281 PP/16 Make DISABLE_DKIM build knob functional.
2283 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2284 Patch by Simon Arlott
2286 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2287 variable argument macros. Our copy now differs from upstream.
2293 TF/01 Failure to get a lock on a hints database can have serious
2294 consequences so log it to the panic log.
2296 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2297 controlled using the smtp_confirmation log selector.
2299 TF/03 Include the error message when we fail to unlink a spool file.
2301 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2302 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2303 for maintaining out-of-tree patches for some time.
2305 PP/01 Bugzilla 139: Documentation and portability issues.
2306 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2307 Handle per-OS dynamic-module compilation flags.
2309 PP/02 Let /dev/null have normal permissions.
2310 The 4.73 fixes were a little too stringent and complained about the
2311 permissions on /dev/null. Exempt it from some checks.
2312 Reported by Andreas M. Kirchwitz.
2314 PP/03 Report version information for many libraries, including
2315 Exim version information for dynamically loaded libraries. Created
2316 version.h, now support a version extension string for distributors
2317 who patch heavily. Dynamic module ABI change.
2319 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2320 privilege escalation vulnerability whereby the Exim run-time user
2321 can cause root to append content of the attacker's choosing to
2324 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2327 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2328 If dropping privileges for untrusted macros, we disabled normal logging
2329 on the basis that it would fail; for the Exim run-time user, this is not
2330 the case, and it resulted in successful deliveries going unlogged.
2331 Fixed. Reported by Andreas Metzler.
2337 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2338 only prepend for the Resent-* case. Fixes regression introduced in
2339 Exim 4.70 by NM/22 for Bugzilla 607.
2341 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2342 increasing numbers of administrators be bitten by this.
2344 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2346 PP/03 Bugzilla 994: added openssl_options main configuration option.
2348 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2350 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2352 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2354 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2355 without a peer certificate, leading to a segfault because of an
2356 assumption that peers always have certificates. Be a little more
2357 paranoid. Problem reported by Martin Tscholak.
2359 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2360 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2361 NB: ClamAV planning to remove STREAM in "middle of 2010".
2362 CL also introduces -bmalware, various -d+acl logging additions and
2363 more caution in buffer sizes.
2365 PP/09 Implemented reverse_ip expansion operator.
2367 PP/10 Bugzilla 937: provide a "debug" ACL control.
2369 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2371 PP/12 Bugzilla 973: Implement --version.
2373 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2375 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2377 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2379 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2380 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2381 ignore trailing whitespace.
2383 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2385 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2388 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2389 Notification from Dr Andrew Aitchison.
2391 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2392 ExtendedDetectionInfo response format.
2393 Notification from John Horne.
2395 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2398 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2399 XSL and documented dependency on system catalogs, with examples of how
2402 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2405 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2406 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2407 configuration file which is writeable by the Exim user or group.
2409 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2410 of configuration files to cover files specified with the -C option if
2411 they are going to be used with root privileges, not just the default
2414 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2415 option (effectively making it always true).
2417 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2418 files to be used while preserving root privileges.
2420 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2421 that rogue child processes cannot use them.
2423 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2424 run-time user, instead of root.
2426 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2427 Exim run-time user without dropping privileges.
2429 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2430 result string, instead of calling string_vformat() twice with the same
2433 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2434 for other users. Others should always drop root privileges if they use
2435 -C on the command line, even for a whitelisted configure file.
2437 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2439 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2445 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2446 $data_path, and $header_path variables; fixed documentation bugs and
2449 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2450 exipick to access non-standard spools, including the "frozen" queue
2453 NM/01 Bugzilla 965: Support mysql stored procedures.
2454 Patch from Alain Williams
2456 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2458 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2459 Patch from Andreas Metzler
2461 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2462 Patch from Kirill Miazine
2464 NM/05 Bugzilla 671: Added umask to procmail example.
2466 JJ/03 installed exipick 20100323.0, fixing doc bug
2468 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2469 directory. Notification and patch from Dan Rosenberg.
2471 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2473 TK/02 Improve log output when DKIM signing operation fails.
2475 MH/01 Treat the transport option dkim_domain as a colon separated
2476 list, not as a single string, and sign the message with each element,
2477 omitting multiple occurences of the same signer.
2479 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2480 Bugzilla 985, 986. Patch by Simon Arlott
2482 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2483 Patch by Simon Arlott
2485 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2486 MBX locking. Notification from Dan Rosenberg.
2492 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2494 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2496 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2498 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2500 NM/04 Bugzilla 915: Flag broken perl installation during build.
2506 TK/01 Added patch by Johannes Berg that expands the main option
2507 "spamd_address" if it starts with a dollar sign.
2509 TK/02 Write list of recipients to X-Envelope-Sender header when building
2510 the mbox-format spool file for content scanning (suggested by Jakob
2513 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2514 (http://www.dcc-servers.net/) support via dccifd. Activated by
2515 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2517 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2518 by Mark Daniel Reidel <mr@df.eu>.
2520 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2521 When building exim an external PCRE library is now needed -
2522 PCRE is a system library on the majority of modern systems.
2523 See entry on PCRE_LIBS in EDITME file.
2525 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2526 conversation. Added nologin parameter to request.
2527 Patch contributed by Kirill Miazine.
2529 TF/01 Do not log submission mode rewrites if they do not change the address.
2531 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2533 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2534 log files in place. Contributed by Roberto Lima.
2536 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2538 TF/03 Bugzilla 615: When checking the local_parts router precondition
2539 after a local_part_suffix or local_part_prefix option, Exim now
2540 does not use the address's named list lookup cache, since this
2541 contains cached lookups for the whole local part.
2543 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2544 Robert Millan. Documentation is in experimental-spec.txt.
2546 TF/04 Bugzilla 668: Fix parallel build (make -j).
2548 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2550 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2551 Patch provided by Jan Srzednicki.
2553 TF/05 Leading white space used to be stripped from $spam_report which
2554 wrecked the formatting. Now it is preserved.
2556 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2557 that they are available at delivery time.
2559 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2561 TF/08 TLS error reporting now respects the incoming_interface and
2562 incoming_port log selectors.
2564 TF/09 Produce a more useful error message if an SMTP transport's hosts
2565 setting expands to an empty string.
2567 NM/06 Bugzilla 744: EXPN did not work under TLS.
2568 Patch provided by Phil Pennock.
2570 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2571 Patch provided by Richard Godbee.
2573 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2574 acl_smtp_notquit, added index entry.
2576 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2577 Patch provided by Eugene Bujak.
2579 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2580 accept(). Patch provided by Maxim Dounin.
2582 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2583 Patch provided by Phil Pennock.
2585 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2587 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2588 Patch provided by Brad "anomie" Jorsch.
2590 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2591 Patch provided by Dean Brooks.
2593 TK/05 Add native DKIM support (does not depend on external libraries).
2595 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2596 Patch provided by Graeme Fowler.
2598 NM/16 Bugzilla 851: Documentation example syntax fix.
2600 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2602 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2605 NM/19 Bugzilla 745: TLS version reporting.
2606 Patch provided by Phil Pennock.
2608 NM/20 Bugzilla 167: bool: condition support.
2609 Patch provided by Phil Pennock.
2611 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2612 clients. Patch provided by Phil Pennock.
2614 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2615 Patch provided by Brad "anomie" Jorsch.
2617 NM/23 Bugzilla 687: Fix misparses in eximstats.
2618 Patch provided by Heiko Schlittermann.
2620 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2621 Patch provided by Heiko Schlittermann.
2623 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2624 plus update to original patch.
2626 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2628 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2629 Patch provided by David Brownlee.
2631 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2633 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2635 NM/30 Bugzilla 888: TLS documentation bugfixes.
2637 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2639 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2640 Unlike the original bugzilla I have changed all shell scripts in src tree.
2642 NM/33 Bugzilla 898: Transport filter timeout fix.
2643 Patch by Todd Rinaldo.
2645 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2646 Patch by Serge Demonchaux.
2648 NM/35 Bugzilla 39: Base64 decode bug fixes.
2649 Patch by Jakob Hirsch.
2651 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2653 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2655 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2657 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2663 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2664 ALT-N's libdkim that I have put here:
2665 http://duncanthrax.net/exim-experimental/
2667 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2668 were called 'true' and it seems that C99 defines that as a reserved
2669 keyword to be used with 'bool' variable types. That means you could
2670 not include C99-style headers which use bools without triggering
2671 build errors in sieve.c.
2673 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2674 as mailq or other aliases. Changed the --help handling significantly
2675 to do whats expected. exim_usage() emits usage/help information.
2677 SC/01 Added the -bylocaldomain option to eximstats.
2679 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2681 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2683 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2689 PH/01 Another patch from the Sieve maintainer.
2691 PH/02 When an IPv6 address is converted to a string for single-key lookup
2692 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2693 dots are used instead of colons so that keys in lsearch files need not
2694 contain colons. This was done some time before quoting was made available
2695 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2696 (notated using the quote facility) so as to distinguish them from IPv4
2697 keys. This meant that lookups for IP addresses in host lists did not work
2698 for iplsearch lookups.
2700 This has been fixed by arranging for IPv6 addresses to be expressed with
2701 colons if the lookup type is iplsearch. This is not incompatible, because
2702 previously such lookups could never work.
2704 The situation is now rather anomalous, since one *can* have colons in
2705 ordinary lsearch keys. However, making the change in all cases is
2706 incompatible and would probably break a number of configurations.
2708 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2711 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2712 conversion specification without a maximum field width, thereby enabling
2713 a rogue spamd server to cause a buffer overflow. While nobody in their
2714 right mind would setup Exim to query an untrusted spamd server, an
2715 attacker that gains access to a server running spamd could potentially
2716 exploit this vulnerability to run arbitrary code as the Exim user.
2718 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2719 $primary_hostname instead of what libspf2 thinks the hosts name is.
2721 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2722 a directory entry by the name of the lookup key. Previously, if a
2723 symlink pointed to a non-existing file or a file in a directory that
2724 Exim lacked permissions to read, a lookup for a key matching that
2725 symlink would fail. Now it is enough that a matching directory entry
2726 exists, symlink or not. (Bugzilla 503.)
2728 PH/03 The body_linecount and body_zerocount variables are now exported in the
2731 PH/04 Added the $dnslist_matched variable.
2733 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2734 This means they are set thereafter only if the connection becomes
2737 PH/06 Added the client_condition to authenticators so that some can be skipped
2738 by clients under certain conditions.
2740 PH/07 The error message for a badly-placed control=no_multiline_responses left
2741 "_responses" off the end of the name.
2743 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2745 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2746 (without spaces) instead of just copying the configuration text.
2748 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2750 PH/11 Added $max_received_linelength.
2752 PH/12 Added +ignore_defer and +include_defer to host lists.
2754 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2755 way in which PCRE > 7.0 is built.
2757 PH/14 Implemented queue_only_load_latch.
2759 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2760 MAIL command. The effect was to mangle the value on 64-bit systems.
2762 PH/16 Another patch from the Sieve maintainer.
2764 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2766 PH/18 If a system quota error occurred while trying to create the file for
2767 a maildir delivery, the message "Mailbox is full" was not appended to the
2768 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2769 only to a quota excession during the actual writing of the file.
2771 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2772 characters?) which causes problems in log lines. The DN values are now
2773 passed through string_printing() before being added to log lines.
2775 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2776 and InterBase are left for another time.)
2778 PH/21 Added message_body_newlines option.
2780 PH/22 Guard against possible overflow in moan_check_errorcopy().
2782 PH/23 POSIX allows open() to be a macro; guard against that.
2784 PH/24 If the recipient of an error message contained an @ in the local part
2785 (suitably quoted, of course), incorrect values were put in $domain and
2786 $local_part during the evaluation of errors_copy.
2792 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2793 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2796 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2797 issue a MAIL command.
2799 PH/02 In an ACL statement such as
2801 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2803 if a client was not listed at all, or was listed with a value other than
2804 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2805 the condition was not true (as it should be), so access was not denied.
2806 The bug was that the ! inversion was incorrectly passed on to the second
2807 item. This has been fixed.
2809 PH/03 Added additional dnslists conditions == and =& which are different from
2810 = and & when the dns lookup returns more than one IP address.
2812 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2813 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2815 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2816 FSYNC, which compiles an option called disable_fsync that allows for
2817 bypassing fsync(). The documentation is heavily laced with warnings.
2819 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2821 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2822 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2823 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2824 including adding "make clean"; (3) Added -fPIC when compiling the test
2825 dynamically loaded module, to get rid of a warning.
2827 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2828 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2829 = 0s. The bug is otherwise harmless.
2831 PH/07 There was a bug in the dovecot authenticator such that the value of
2832 $auth1 could be overwritten, and so not correctly preserved, after a
2833 successful authentication. This usually meant that the value preserved by
2834 the server_setid option was incorrect.
2836 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2838 PH/09 Installed PCRE release 7.0.
2840 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2841 run for batched SMTP input. It is now run at the start of every message
2842 in the batch. While fixing this I discovered that the process information
2843 (output by running exiwhat) was not always getting set for -bs and -bS
2844 input. This is fixed, and it now also says "batched" for BSMTP.
2846 PH/11 Added control=no_pipelining.
2848 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2849 patch, slightly modified), and move the expansion of helo_data till after
2850 the connection is made in the smtp transport (so it can use these
2853 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2855 PH/14 Added log_selector = +pid.
2857 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2859 PH/16 Add ${if forany and ${if forall.
2861 PH/17 Added dsn_from option to vary the From: line in DSNs.
2863 PH/18 Flush SMTP output before performing a callout, unless control =
2864 no_callout_flush is set.
2866 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2867 was true (the default) a successful delivery failed to delete the retry
2868 item, thus causing premature timeout of the address. The bug is now
2871 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2873 PH/21 Long custom messages for fakedefer and fakereject are now split up
2874 into multiline responses in the same way that messages for "deny" and
2875 other ACL rejections are.
2877 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2878 with slight modification.
2880 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2881 draft, changing the syntax and factoring some duplicate code".
2883 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2884 for deliveries of the second and subsequent messages over the same SMTP
2887 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2888 ${reduce, with only minor "tidies".
2890 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2892 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2893 expansion side effects.
2895 PH/27 When a message times out after an over-quota error from an Exim-imposed
2896 quota, the bounce message says "mailbox is full". This message was not
2897 being given when it was a system quota that was exceeded. It now should
2900 MH/03 Made $recipients available in local_scan(). local_scan() already has
2901 better access to the recipient list through recipients_list[], but
2902 $recipients can be useful in postmaster-provided expansion strings.
2904 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2905 in the case of a MAIL command with additional options following the
2906 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2907 were accidentally chopped off.
2909 PH/29 SMTP synchronization checks are implemented when a command is read -
2910 there is a check that no more input is waiting when there shouldn't be
2911 any. However, for some commands, a delay in an ACL can mean that it is
2912 some time before the response is written. In this time, more input might
2913 arrive, invalidly. So now there are extra checks after an ACL has run for
2914 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2915 pipelining has not been advertised.
2917 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2919 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2920 returned in the SMTP response when smtp_return_error_details was false.
2921 This has been fixed.
2923 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2924 instead of the C I/O that was originally supplied, because problems were
2925 reported on Solaris.
2927 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2928 Exim which did not show up earlier: it was assuming that a call to
2929 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2930 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2931 was a macro that became an assignment, so it seemed to work. This has
2932 changed to a proper function call with a void return, hence the compile
2933 error. Exim's code has been fixed.
2935 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2938 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2940 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2942 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2943 because it turns out to be too big or there is a timeout) there is no
2944 "Completed" line in the log. When some messages of this type were
2945 selected by exigrep, they were listed as "not completed". Others were
2946 picked up by some special patterns. I have improved the selection
2947 criteria to be more general.
2949 PH/38 The host_find_failed option in the manualroute router can now be set
2950 to "ignore", to completely ignore a host whose IP address cannot be
2951 found. If all hosts are ignored, the behaviour is controlled by the new
2952 host_all_ignored option.
2954 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2955 homing or because of multiple MX records with /mx) generated more than
2956 one IP address, and the following item turned out to be the local host,
2957 all the secondary addresses of the first item were incorrectly removed
2958 from the list, along with the local host and any following hosts (which
2959 is what is supposed to happen).
2961 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2962 whoever called Exim into the -H file. In the case of the daemon it was
2963 behaving confusingly. When first started, it used values for whoever
2964 started the daemon, but after a SIGHUP it used the Exim user (because it
2965 calls itself on a restart). I have changed the code so that it now always
2968 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2969 message are rejected with the same error (e.g. no authentication or bad
2970 sender address), and a DATA command is nevertheless sent (as can happen
2971 with PIPELINING or a stupid MUA), the error message that was given to the
2972 RCPT commands is included in the rejection of the DATA command. This is
2973 intended to be helpful for MUAs that show only the final error to their
2976 PH/42 Another patch from the Sieve maintainer.
2978 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2979 Eximstats - Fixed some broken HTML links and added missing column headers
2981 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2982 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2984 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2986 PH/43 Yet another patch from the Sieve maintainer.
2988 PH/44 I found a way to check for a TCP/IP connection going away before sending
2989 the response to the final '.' that terminates a message, but only in the
2990 case where the client has not sent further data following the '.'
2991 (unfortunately, this is allowed). However, in many cases there won't be
2992 any further data because there won't be any more messages to send. A call
2993 to select() can be used: if it shows that the input is "ready", there is
2994 either input waiting, or the socket has been closed. An attempt to read
2995 the next input character can distinguish the two cases. Previously, Exim
2996 would have sent an OK response which the client would never have see.
2997 This could lead to message repetition. This fix should cure that, at
2998 least in a lot of common cases.
3000 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3001 advertised in response to EHLO.
3007 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3008 fixed by 4.65/MH/01 (is this a record?) are fixed:
3010 (i) An empty string was always treated as zero by the numeric comparison
3011 operators. This behaviour has been restored.
3013 (ii) It is documented that the numeric comparison operators always treat
3014 their arguments as decimal numbers. This was broken in that numbers
3015 starting with 0 were being interpreted as octal.
3017 While fixing these problems I realized that there was another issue that
3018 hadn't been noticed. Values of message_size_limit (both the global option
3019 and the transport option) were treated as octal if they started with 0.
3020 The documentation was vague. These values are now always treated as
3021 decimal, and I will make that clear in the documentation.
3027 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3028 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3031 MH/01 Don't check that the operands of numeric comparison operators are
3032 integers when their expansion is in "skipping" mode (fixes bug
3033 introduced by 4.64-PH/07).
3035 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3036 child addresses, Exim now panics and dies. Previously, because the count
3037 is held in a short int, deliveries were likely to be lost. As such a
3038 large number of recipients for a single message is ridiculous
3039 (performance will be very, very poor), I have chosen to impose a limit
3040 rather than extend the field.
3046 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3047 leftover -K file (the existence of which was triggered by #402).
3048 While we were at it, introduced process PID as part of the -K
3049 filename. This should rule out race conditions when creating
3052 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3053 processing considerably. Previous code took too long for large mails,
3054 triggering a timeout which in turn triggers #401.
3056 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3057 in the DK code in transports.c. sendfile() is not really portable,
3058 hence the _LINUX specificness.
3060 TF/01 In the add_headers option to the mail command in an Exim filter,
3061 there was a bug that Exim would claim a syntax error in any
3062 header after the first one which had an odd number of characters
3065 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3066 callout verification, Exim cached a "reject" for the entire domain. This
3067 is correct for most verifications, but it is not correct for a recipient
3068 verification with use_sender or use_postmaster set, because in that case
3069 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3070 case of MAIL FROM:<> rejection from other early rejections (e.g.
3071 rejection of HELO). When verifying a recipient using a non-null MAIL
3072 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3073 Whatever the result of the callout, the value of the domain cache is
3074 left unchanged (for any other kind of callout, getting as far as trying
3075 RCPT means that the domain itself is ok).
3077 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3080 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3081 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3082 session, the daemon ignores ECONNECT errors and logs others; it now
3083 ignores EPIPE as well.
3085 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3086 (quoted-printable decoding).
3088 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3089 later the small subsequent patch to fix an introduced bug.
3091 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3093 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3095 PH/08 An error is now given if message_size_limit is specified negative.
3097 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3098 to be given (somewhat) arbitrary names.
3100 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3103 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3104 miscellaneous code fixes
3106 PH/10 Added the log_reject_target ACL modifier to specify where to log
3109 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3110 hostname. This is wrong, because it relates to the incoming message (and
3111 probably the interface on which it is arriving) and not to the outgoing
3112 callout (which could be using a different interface). This has been
3113 changed to use the value of the helo_data option from the smtp transport
3114 instead - this is what is used when a message is actually being sent. If
3115 there is no remote transport (possible with a router that sets up host
3116 addresses), $smtp_active_hostname is used.
3118 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3119 tweaks were necessary in order to get it to work (see also 21 below):
3120 (a) The code assumed that strncpy() returns a negative number on buffer
3121 overflow, which isn't the case. Replaced with Exim's string_format()
3123 (b) There were several signed/unsigned issues. I just did the minimum
3124 hacking in of casts. There is scope for a larger refactoring.
3125 (c) The code used strcasecmp() which is not a standard C function.
3126 Replaced with Exim's strcmpic() function.
3127 (d) The code set only $1; it now sets $auth1 as well.
3128 (e) A simple test gave the error "authentication client didn't specify
3129 service in request". It would seem that Dovecot has changed its
3130 interface. Fortunately there's a specification; I followed it and
3131 changed what the client sends and it appears to be working now.
3133 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3136 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3137 address A is aliased to B and C, where B exists and C does not. Without
3138 -v the output is "A verified" because verification stops after a
3139 successful redirection if more than one address is generated. However,
3140 with -v the child addresses are also verified. Exim was outputting "A
3141 failed to verify" and then showing the successful verification for C,
3142 with its parentage. It now outputs "B failed to verify", showing B's
3143 parentage before showing the successful verification of C.
3145 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3146 look up a TXT record in a specific list after matching in a combined
3149 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3150 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3151 they consult the DNS. I had assumed they would set it the way they
3152 wanted; and indeed my experiments on Linux seem to show that in some
3153 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3154 To be on the safe side, however, I have now made the interface to
3155 host_find_byname() similar to host_find_bydns(), with an argument
3156 containing the DNS resolver options. The host_find_byname() function now
3157 sets these options at its start, just as host_find_bydns() does. The smtp
3158 transport options dns_qualify_single and dns_search_parents are passed to
3159 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3160 of host_find_byname() use the default settings of RES_DEFNAMES
3161 (qualify_single) but not RES_DNSRCH (search_parents).
3163 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3164 spool_read_header() do less string testing, by means of a preliminary
3165 switch on the second character of optional "-foo" lines. (This is
3166 overdue, caused by the large number of possibilities that now exist.
3167 Originally there were few.) While I was there, I also converted the
3168 str(n)cmp tests so they don't re-test the leading "-" and the first
3169 character, in the hope this might squeeze out yet more improvement.
3171 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3172 flag allowing group syntax was set by the header_syntax check but not
3173 turned off, possible causing trouble later; (2) The flag was not being
3174 set at all for the header_verify test, causing "group"-style headers to
3175 be rejected. I have now set it in this case, and also caused header_
3176 verify to ignore an empty address taken from a group. While doing this, I
3177 came across some other cases where the code for allowing group syntax
3178 while scanning a header line wasn't quite right (mostly, not resetting
3179 the flag correctly in the right place). These bugs could have caused
3180 trouble for malformed header lines. I hope it is now all correct.
3182 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3183 with the "reply" argument non-NULL. The code, however (which originally
3184 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3185 but it didn't always do it. This confused somebody who was copying the
3186 code for some other use. I have removed all the tests.
3188 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3189 feature that was used to support insecure browsers during the U.S. crypto
3190 embargo. It requires special client support, and Exim is probably the
3191 only MTA that supported it -- and would never use it because real RSA is
3192 always available. This code has been removed, because it had the bad
3193 effect of slowing Exim down by computing (never used) parameters for the
3194 RSA_EXPORT functionality.
3196 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3197 authenticator to fail if there's a tab character in the incoming data
3198 (there should never be unless someone is messing about, as it's supposed
3199 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3200 if the connection is using TLS or if the remote IP is the same as the
3201 local IP, and the "valid-client-cert option" if a client certificate has
3204 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3205 authenticators. This can be used for authorization after authentication
3206 succeeds. (In the case of plaintext, it servers for both authentication
3209 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3210 if any retry times were supplied.
3212 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3213 connection, where there is no client IP address to check. In this
3214 situation, the verify now always succeeds.
3216 PH/25 Applied John Jetmore's -Mset patch.
3218 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3220 PH/27 In a string expansion for a processed (not raw) header when multiple
3221 headers of the same name were present, leading whitespace was being
3222 removed from all of them, but trailing whitespace was being removed only
3223 from the last one. Now trailing whitespace is removed from each header
3224 before concatenation. Completely empty headers in a concatenation (as
3225 before) are ignored.
3227 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3228 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3230 PH/29 [Removed. This was a change that I later backed out, and forgot to
3231 correct the ChangeLog entry (that I had efficiently created) before
3232 committing the later change.]
3234 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3235 address errors (4xx response to RCPT) over the same connection as other
3236 messages routed to the same hosts. Such deliveries are always "forced",
3237 so retry times are not inspected. This resulted in far too many retries
3238 for the affected addresses. The effect occurred only when there were more
3239 hosts than the hosts_max_try setting in the smtp transport when it had
3240 the 4xx errors. Those hosts that it had tried were not added to the list
3241 of hosts for which the message was waiting, so if all were tried, there
3242 was no problem. Two fixes have been applied:
3244 (i) If there are any address or message errors in an SMTP delivery, none
3245 of the hosts (tried or untried) are now added to the list of hosts
3246 for which the message is waiting, so the message should not be a
3247 candidate for sending over the same connection that was used for a
3248 successful delivery of some other message. This seems entirely
3249 reasonable: after all the message is NOT "waiting for some host".
3250 This is so "obvious" that I'm not sure why it wasn't done
3251 previously. Hope I haven't missed anything, but it can't do any
3252 harm, as the worst effect is to miss an optimization.
3254 (ii) If, despite (i), such a delivery is accidentally attempted, the
3255 routing retry time is respected, so at least it doesn't keep
3256 hammering the server.
3258 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3259 in ${readsocket because some servers need this prod.
3261 PH/32 Added some extra debug output when updating a wait-xxx database.
3263 PH/33 The hint "could be header name not terminated by colon", which has been
3264 given for certain expansion errors for a long time, was not being given
3265 for the ${if def:h_colon_omitted{... case.
3267 PH/34 The spec says: "With one important exception, whenever a domain list is
3268 being scanned, $domain contains the subject domain." There was at least
3269 one case where this was not true.
3271 PH/35 The error "getsockname() failed: connection reset by peer" was being
3272 written to the panic log as well as the main log, but it isn't really
3273 panic-worthy as it just means the connection died rather early on. I have
3274 removed the panic log writing for the ECONNRESET error when getsockname()
3277 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3278 runs only) independently of the message's sender address. This meant
3279 that, if the 4xx error was in fact related to the sender, a different
3280 message to the same recipient with a different sender could confuse
3281 things. In particular, this can happen when sending to a greylisting
3282 server, but other circumstances could also provoke similar problems.
3283 I have changed the default so that the retry time for these errors is now
3284 based a combination of the sender and recipient addresses. This change
3285 can be overridden by setting address_retry_include_sender=false in the
3288 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3289 remote server are returned as part of bounce messages. This was not
3290 happening for LMTP over a pipe (the lmtp transport), but now it is the
3291 same for both kinds of LMTP.
3293 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3294 in header lines that were in fact CNAMEs. This is no longer the case.
3296 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3297 and queue runs started by the daemon processed all messages. This has
3298 been fixed so that -R and -S can now usefully be given with -q<time>.
3300 PH/40 Import PCRE release 6.7 (fixes some bugs).
3302 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3304 PH/42 Give an error if -q is specified more than once.
3306 PH/43 Renamed the variables $interface_address and $interface_port as
3307 $received_ip_address and $received_port, to make it clear that these
3308 values apply to message reception, and not to the outgoing interface when
3309 a message is delivered. (The old names remain recognized, of course.)
3311 PH/44 There was no timeout on the connect() call when using a Unix domain
3312 socket in the ${readsocket expansion. There now is.
3314 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3315 be meaningful with "accept".
3317 SC/01 Eximstats V1.43
3318 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3320 SC/02 Eximstats V1.44
3321 Use a glob alias rather than an array ref in the generated
3322 parser. This improves both readability and performance.
3324 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3325 Collect SpamAssassin and rejection statistics.
3326 Don't display local sender or destination tables unless
3327 there is data to show.
3328 Added average volumes into the top table text output.
3330 SC/04 Eximstats V1.46
3331 Collect data on the number of addresses (recipients)
3332 as well as the number of messages.
3334 SC/05 Eximstats V1.47
3335 Added 'Message too big' to the list of mail rejection
3336 reasons (thanks to Marco Gaiarin).
3338 SC/06 Eximstats V1.48
3339 Mainlog lines which have GMT offsets and are too short to
3340 have a flag are now skipped.
3342 SC/07 Eximstats V1.49 (Alain Williams)
3343 Added the -emptyok flag.
3345 SC/08 Eximstats V1.50
3346 Fixes for obtaining the IP address from reject messages.
3348 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3349 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3350 whitespace changes from 4.64-PH/27
3352 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3355 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3356 are found, allow negative numbers in numeric criteria)
3358 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3360 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3361 to match changes made in 4.64-PH/43
3363 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3365 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3366 "domain" is actually an IP address, and force a failure. This locks out
3367 those revolvers/nameservers that support "A-for-A" lookups, in
3368 contravention of the specifications.
3370 PH/48 When a host name was looked up from an IP address, and the subsequent
3371 forward lookup of the name timed out, the host name was left in
3372 $sender_host_name, contrary to the specification.
3374 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3375 restricted to single-key lookups, Exim was not diagnosing an error if
3376 * or *@ was used with a query-style lookup.
3378 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3380 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3381 long ago, but noone interested enough thought of it. Let's just say that
3382 the "1.1" means that there are some new functions that weren't there at
3383 some point in the past.
3385 PH/51 Error processing for expansion failure of helo_data from an smtp
3386 transport during callout processing was broken.
3388 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3389 tested/used via the -bh/-bhc/-bs options.
3391 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3392 bug, fixed in subsequent PCRE releases).
3394 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3395 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3397 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3403 SC/01 Use a glob alias rather than an array ref in eximstats generated
3404 parser. This improves both readability and performance.
3406 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3407 Don't display local sender or destination tables in eximstats unless
3408 there is data to show.
3409 Added average volumes into the eximstats top table text output.
3411 SC/03 Collect data on the number of addresses (recipients) as well
3412 as the number of messages in eximstats.
3414 TF/01 Correct an error in the documentation for the redirect router. Exim
3415 does (usually) call initgroups() when daemonizing.
3417 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3418 with consistent privilege compared to when running as a daemon.
3420 TF/03 Note in the spec that $authenticated_id is not set for local
3421 submissions from trusted users.
3423 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3424 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3426 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3427 by adding some example configuration directives to the default
3428 configuration file. A little bit of work is required to uncomment the
3429 directives and define how usernames and passwords are checked, but
3430 there is now a framework to start from.
3432 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3433 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3434 without this. I don't know how relevant this is to other LDAP libraries.
3436 PH/02 Add the verb name to the "unknown ACL verb" error.
3438 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3440 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3442 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3443 directory not expanded when it should be if an expanded home directory
3444 was set for the address (which is overridden by the transport).
3446 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3449 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3450 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3451 because it is too late at that time, and has no effect.
3453 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3454 security issue with \' (bugzilla #107). I could not use the
3455 PQescapeStringConn() function, because it needs a PGconn value as one of
3458 PH/08 When testing addresses using -bt, indicate those final addresses that
3459 are duplicates that would not cause an additional delivery. At least one
3460 person was confused, thinking that -bt output corresponded to deliveries.
3461 (Suppressing duplicates isn't a good idea as you lose the information
3462 about possibly different redirections that led to the duplicates.)
3464 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3465 systems where poll() doesn't work, in particular OS X.
3467 PH/10 Added more information to debugging output for retry time not reached.
3469 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3470 operations in malware.c.
3472 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3475 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3476 syslog_duplication set false, log lines that would normally be written
3477 both the the main log and to the reject log were not written to syslog at
3480 PH/14 In the default configuration, change the use of "message" in ACL warn
3481 statements to "add_header".
3483 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3484 not followed by a command (e.g. "seen endif").
3486 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3487 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3490 PH/17 Added extra conditions to the default value of delay_warning_condition
3494 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3495 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3496 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3499 The Auto-Submitted: and various List- headers are standardised, whereas I
3500 don't think Precedence: ever was.
3502 PH/18 Refactored debugging code in route_finduser() to show more information,
3503 in particular, the error code if getpwnam() issues one.
3505 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3506 This is apparently needed in addition to the PH/07 change above to avoid
3507 any possible encoding problems.
3509 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3510 but not after initializing Perl.
3512 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3513 output them only if debugging. By default they are written stderr,
3514 apparently, which is not desirable.
3516 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3519 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3522 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3524 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3525 authenticated or an ident call has been made. Suppress the default
3526 values for $authenticated_id and $authenticated_sender (but permit -oMai
3527 and -oMas) when testing with -bh.
3529 PH/24 Re-jigged the order of the tests in the default configuration so that the
3530 tests for valid domains and recipients precede the DNS black list and CSA
3531 tests, on the grounds that those ones are more expensive.
3533 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3534 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3535 This bug exists in every version of Exim that I still have, right back to
3538 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3539 However, an attempt to turn on case-sensitivity in a regex key by
3540 including (?-i) didn't work because the subject string was already
3541 lowercased, and the effects were non-intuitive. It turns out that a
3542 one-line patch can be used to allow (?-i) to work as expected.
3548 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3549 other effects) broke the use of negated acl sub-conditions.
3551 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3554 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3555 "Deny" causes Exim to reject the incoming connection with a 554 error.
3556 Unfortunately, if there is a major crisis, such as a disk failure,
3557 tcp-wrappers gives "deny", whereas what one would like would be some
3558 kind of temporary error. A kludge has been added to help with this.
3559 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3560 554 error is used if errno is still zero or contains ENOENT (which occurs
3561 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3564 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3566 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3567 errors. Otherwise a message that provokes a temporary error (when other
3568 messages do not) can cause a whole host to time out.
3570 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3571 addresses were routed directly to files or pipes from a redirect router.
3572 File deliveries just didn't batch; pipe deliveries might have suffered
3575 PH/06 A failure to get a lock for a hints database would erroneously always say
3576 "Failed to get write lock", even when it was really a read lock.
3578 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3579 of 0600. This has been changed to use the value of the lockfile_mode
3580 option (which defaults to 0600).
3582 PH/08 Applied small patch from the Sieve maintainer.
3584 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3585 folder from quota calculations, a direct delivery into this folder messed
3586 up the contents of the maildirsize file. This was because the regex was
3587 used only to exclude .Trash (or whatever) when the size of the mailbox
3588 was calculated. There was no check that a delivery was happening into an
3589 excluded directory. This bug has been fixed by ignoring all quota
3590 processing for deliveries into excluded directories.
3592 PH/10 Added the maildirfolder_create_regex option to appendfile.
3598 PH/01 The code for finding all the local interface addresses on a FreeBSD
3599 system running IPv6 was broken. This may well have applied to all BSD
3600 systems, as well as to others that have similar system calls. The broken
3601 code found IPv4 interfaces correctly, but gave incorrect values for the
3602 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3603 that it would not match correctly against @[] and not recognize the IPv6
3606 PH/02 The ipliteral router was not recognizing addresses of the form user@
3607 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3609 PH/03 Added disable_ipv6.
3611 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3612 decoded form, because it is most often used to construct To: headers
3613 lines in autoreplies, and the decoded form may well be syntactically
3614 invalid. However, $reply_address has leading white space removed, and all
3615 newlines turned into spaces so that the autoreply transport does not
3618 PH/05 If group was specified without a user on a router, and no group or user
3619 was specified on a transport, the group from the router was ignored.
3621 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3622 for visible compile-time settings that can be used to change these
3623 numbers, for those that want even more. Backwards compatibility with old
3624 spool files has been maintained. However, going back to a previous Exim
3625 release will lost any variables that are in spool files.
3627 PH/07 Two small changes when running in the test harness: increase delay when
3628 passing a TCP/IP connection to a new process, in case the original
3629 process has to generate a bounce, and remove special handling of
3630 127.0.0.2 (sic), which is no longer necessary.
3632 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3633 be the same on different OS.
3635 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3638 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3639 whether --show-vars was specified or not
3641 JJ/02 exipick: Added support for new ACL variable spool format introduced
3644 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3645 syntactically invalid From: or Reply-to: line, and a filter used this to
3646 generate an autoreply, and therefore failed to obtain an address for the
3647 autoreply, Exim could try to deliver to a non-existent relative file
3648 name, causing unrelated and misleading errors. What now happens is that
3649 it logs this as a hard delivery error, but does not attempt to create a
3652 PH/11 The exinext utility has a -C option for testing purposes, but although
3653 the given file was scanned by exinext itself; it wasn't being passed on
3654 when Exim was called.
3656 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3657 an end-of-file indication when reading a command response.
3659 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3660 compiled. In many other places in Exim, IPv6 addresses are always
3661 recognized, so I have changed this. It also means that IPv4 domain
3662 literals of the form [IPV4:n.n.n.n] are now always recognized.
3664 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3665 used if the router is not running as root, for example, when verifying at
3666 ACL time, or when using -bh. The debugging output from this situation was
3667 non-existent - all you got was a failure to exec. I have made two
3670 (a) Failures to set uid/gid, the current directory, or a process leader
3671 in a subprocess such as that created by queryprogram now generate
3672 suitable debugging output when -d is set.
3674 (b) The queryprogram router detects when it is not running as root,
3675 outputs suitable debugging information if -d is set, and then runs
3676 the subprocess without attempting to change uid/gid.
3678 PH/15 Minor change to Makefile for building test_host (undocumented testing
3681 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3682 additional section of a DNS packet that returns MX or SRV records.
3683 Instead, it always explicitly searches for A/AAAA records. This avoids
3684 major problems that occur when a DNS server includes only records of one
3685 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3686 fixed another bug: if SRV records were looked up and the corresponding
3687 address records were *not* found in the additional section, the port
3688 values from the SRV records were lost.
3690 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3691 using the correct key (the original address) when searching the retry
3692 rules in order to find which one to use for generating the retry hint.
3694 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3695 adding the default one. Similarly, if it contains a Reply-To: header, the
3696 errors_reply_to option, if set, is not used.
3698 PH/19 When calculating a retry time, Exim used to measure the "time since
3699 failure" by looking at the "first failed" field in the retry record. Now
3700 it does not use this if it is later than than the arrival time of the
3701 message. Instead it uses the arrival time. This makes for better
3702 behaviour in cases where some deliveries succeed, thus re-setting the
3703 "first failed" field. An example is a quota failure for a huge message
3704 when small messages continue to be delivered. Without this change, the
3705 "time since failure" will always be short, possible causing more frequent
3706 delivery attempts for the huge message than are intended.
3707 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3709 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3710 $1, $2, $3) because the numerical variables can be reset during some
3711 expansion items (e.g. "match"), thereby losing the authentication data.
3713 PH/21 Make -bV show the size of off_t variables so that the test suite can
3714 decide whether to run tests for quotas > 2G.
3716 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3717 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3718 filecount value is greater than 2G or if a quota value is greater than 2G
3719 on a system where the size of off_t is not greater than 4, a panic error
3722 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3723 never match. The debug and -bh output now contains an explicit error
3724 message indicating a malformed IPv4 address or mask.
3726 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3727 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3728 PH/23 above applies.
3730 PH/25 Do not write to syslog when running in the test harness. The only
3731 occasion when this arises is a failure to open the main or panic logs
3732 (for which there is an explicit test).
3734 PH/26 Added the /no_tell option to "control=freeze".
3736 PH/27 If a host name lookup failed very early in a connection, for example, if
3737 the IP address matched host_lookup and the reverse lookup yielded a name
3738 that did not have a forward lookup, an error message of the form "no IP
3739 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3740 could be logged. Now it outputs the IP address instead of "NULL".
3742 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3743 allows the sender and the authenticated sender to be set when
3744 submitting a message from within Exim. Since child_open_exim() is
3745 documented for local_scan(), the new function should be too.
3747 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3748 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3749 results in an empty string is now treated as unset.
3751 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3753 PH/31 Added sender_verify_fail logging option.
3755 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3756 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3757 tidied the source and removed it altogether.
3759 PH/33 When a queue run was abandoned because the load average was too high, a
3760 log line was always written; now it is written only if the queue_run log
3761 selector is set. In addition, the log line for abandonment now contains
3762 information about the queue run such as the pid. This is always present
3763 in "start" and "stop" lines but was omitted from the "abandon" line.
3765 PH/34 Omit spaces between a header name and the colon in the error message that
3766 is given when verify = headers_syntax fails (if there are lots of them,
3767 the message gets confusing).
3769 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3770 names, as there are now some PTR records that contain slashes. This check
3771 is only to protect against broken name servers that fall over on strange
3772 characters, so the fact that it applies to all lookups doesn't matter.
3774 PH/36 Now that the new test suite is complete, we can remove some of the
3775 special code in Exim that was needed for the old test suite. For example,
3776 sorting DNS records because real resolvers return them in an arbitrary
3777 order. The new test suite's fake resolver always returns records in the
3780 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3781 bounces) except when queue_only is set, to avoid logging races between
3782 the different processes.
3784 PH/38 Panic-die if .include specifies a non-absolute path.
3786 PH/39 A tweak to the "H" retry rule from its user.
3788 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3789 a label. They prevented compilation on older perls.
3791 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3792 a warning to be raised on newish perls.
3794 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3795 on queue. Changes to match documented behaviour of showing count of
3796 messages matching specified criteria.
3798 PH/40 Changed the default ident timeout from 30s to 5s.
3800 PH/41 Added support for the use of login_cap features, on those BSD systems
3801 that have them, for controlling the resources used by pipe deliveries.
3803 PH/42 The content-scanning code uses fopen() to create files in which to put
3804 message data. Previously it was not paying any attention to the mode of
3805 the files. Exim runs with umask(0) because the rest of the code creates
3806 files with open(), and sets the required mode explicitly. Thus, these
3807 files were ending up world-writeable. This was not a big issue, because,
3808 being within the spool directory, they were not world-accessible. I have
3809 created a function called modefopen, which takes an additional mode
3810 argument. It sets umask(777), creates the file, chmods it to the required
3811 mode, then resets the umask. All the relevant calls to fopen() in the
3812 content scanning code have been changed to use this function.
3814 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3815 to 24 hours. This avoids potential overflow problems when processing G
3816 and H retry rules. I suspect nobody ever tinkers with this value.
3818 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3820 PH/45 When the plaintext authenticator is running as a client, the server's
3821 challenges are checked to ensure they are valid base64 strings. By
3822 default, the authentication attempt is cancelled if an invalid string is
3823 received. Setting client_ignore_invalid_base64 true ignores these errors.
3824 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3825 they are received. Thus, the responses can be made to depend on the
3826 challenges. If an invalid string is ignored, an empty string is placed in
3829 PH/46 Messages that are created by the autoreply transport now contains a
3830 References: header, in accordance with RFCs 2822 and 3834.
3832 PH/47 Added authenticated_sender_force to the smtp transport.
3834 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3836 PH/49 Installed latest patch from the Sieve maintainer.
3838 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3839 was also set, the appendfile transport was unnecessarily scanning a
3840 directory of message files (e.g. for maildir delivery) to find the count
3841 of files (along with the size), even though it did not need this
3842 information. It now does the scan only if it needs to find either the
3843 size of the count of files.
3845 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3847 PH/52 Two bugs concerned with error handling when the smtp transport is
3850 (i) Exim was not creating retry information for temporary errors given
3851 for individual recipients after the DATA command when the smtp transport
3852 was used in LMTP mode. This meant that they could be retried too
3853 frequently, and not timed out correctly.
3855 (ii) Exim was setting the flag that allows error details to be returned
3856 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3857 recipients that were returned after the DATA command.
3859 PH/53 This is related to PH/52, but is more general: for any failing address,
3860 when detailed error information was permitted to be returned to the
3861 sender, but the error was temporary, then after the final timeout, only
3862 "retry timeout exceeded" was returned. Now it returns the full error as
3863 well as "retry timeout exceeded".
3865 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3866 do this, and (what is worse) MTAs that accept it.
3868 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3869 will now be deprecated.
3871 PH/56 New os.c-cygwin from the Cygwin maintainer.
3873 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3874 formats (previously only available in exim formats via -bpr, -bpru,
3875 and -bpra. Now also available in native and exiqgrep formats)
3877 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3878 with very large, slow to parse queues
3880 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3882 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3884 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3885 responses to authentication challenges, though it was showing the
3886 challenges; (ii) I've removed the CR characters from the debug output for
3889 PH/58 Allow for the insertion of a newline as well as a space when a string
3890 is turned into more than one encoded-word during RFC 2047 encoding. The
3891 Sieve code now uses this.
3893 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3894 data_4xx, lost_connection, tls_required.
3896 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3897 message was being sent as an SMTP response.
3899 PH/61 Add -l and -k options to exicyclog.
3901 PH/62 When verifying, if an address was redirected to one new address, so that
3902 verification continued, and the new address failed or deferred after
3903 having set something in $address_data, the value of $address_data was not
3904 passed back to the ACL. This was different to the case when no
3905 redirection occurred. The value is now passed back in both cases.
3907 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3908 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3909 use login_cap.h, so on its own it isn't the distinguishing feature. The
3910 new name refers directly to the setclassresources() function.
3912 PH/65 Added configuration files for NetBSD3.
3914 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3916 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3917 is preferred over IPv4.
3919 PH/68 The bounce_return_message and bounce_return_body options were not being
3920 honoured for bounces generated during the reception of non-SMTP messages.
3921 In particular, this applied to messages rejected by the ACL. This bug has
3922 been fixed. However, if bounce_return_message is true and bounce_return_
3923 body is false, the headers that are returned for a non-SMTP message
3924 include only those that have been read before the error was detected.
3925 (In the case of an ACL rejection, they have all been read.)
3927 PH/69 The HTML version of the specification is now built in a directory called
3928 spec_html instead of spec.html, because the latter looks like a path with
3929 a MIME-type, and this confuses some software.
3931 PH/70 Catch two compiler warnings in sieve.c.
3933 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3934 function verify_get_ident() calls ip_connect() to connect a socket, but
3935 if the "connect()" function timed out, ip_connect() used to close the
3936 socket. However, verify_get_ident() also closes the socket later, and in
3937 between Exim writes to the log, which may get opened at this point. When
3938 the socket was closed in ip_connect(), the log could get the same file
3939 descriptor number as the socket. This naturally causes chaos. The fix is
3940 not to close the socket in ip_connect(); the socket should be closed by
3941 the function that creates it. There was only one place in the code where
3942 this was missing, in the iplookup router, which I don't think anybody now
3943 uses, but I've fixed it anyway.
3945 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3946 well as to direct DNS lookups. Otherwise the handling of names in host
3947 lists is inconsistent and therefore confusing.
3953 PH/01 Two changes to the default runtime configuration:
3955 (1) Move the checks for relay_from_hosts and authenticated clients from
3956 after to before the (commented out) DNS black list checks.
3958 (2) Add control=submission to the relay_from_hosts and authenticated
3959 clients checks, on the grounds that messages accepted by these
3960 statements are most likely to be submissions.
3962 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3964 (1) Generate an error if the third argument for the ${prvs expansion is
3967 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3970 (3) Reset the variables that are obtained from the first argument of
3971 ${prvscheck and used in the second argument before leaving the code,
3972 because their memory is reclaimed, so using them afterwards may do
3975 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3976 one (it's much easier than Tom thought :-).
3978 (5) Because of (4), we can now allow for the use of $prvscheck_result
3979 inside the third argument.
3981 PH/03 For some reason, the default setting of PATH when running a command from
3982 a pipe transport was just "/usr/bin". I have changed it to
3985 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3986 anything to be listed in the output from -bV.
3988 PH/05 When a filter generated an autoreply, the entire To: header line was
3989 quoted in the delivery log line, like this:
3991 => >A.N.Other <ano@some.domain> <original@ddress> ...
3993 This has been changed so that it extracts the operative address. There
3994 may be more than one such address. If so, they are comma-separated, like
3997 => >ano@some.domain,ona@other.domain <original@ddress> ...
3999 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4000 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4001 not being looked up in the rDNS to get a host name, Exim was showing the
4002 IP address twice in Received: lines, even though the IP addresses were
4003 identical. For example:
4005 Received: from [1.2.3.4] (helo=[1.2.3.4])
4007 However, if the real host name was known, it was omitting the HELO data
4008 if it matched the actual IP address. This has been tidied up so that it
4009 doesn't show the same IP address twice.
4011 PH/07 When both +timestamp and +memory debugging was on, the value given by
4012 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4013 called by the debug printing, thereby overwriting the timestamp buffer.
4014 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4016 PH/08 When the original message was included in an autoreply transport, it
4017 always said "this is a copy of the message, including all the headers",
4018 even if body_only or headers_only was set. It now gives an appropriate
4021 PH/09 Applied a patch from the Sieve maintainer which:
4023 o fixes some comments
4024 o adds the (disabled) notify extension core
4025 o adds some debug output for the result of if/elsif tests
4026 o points to the current vacation draft in the documentation
4027 and documents the missing references header update
4031 o fixes a bug in processing the envelope test (when testing
4032 multiple envelope elements, the last element determined the
4035 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4036 Electronic Mail") by including:
4038 Auto-submitted: auto-generated
4040 in the messages that it generates (bounce messages and others, such as
4041 warnings). In the case of bounce messages for non-SMTP messages, there was
4042 also a typo: it was using "Auto_submitted" (underscore instead of
4043 hyphen). Since every message generated by Exim is necessarily in response
4044 to another message, thes have all been changed to:
4046 Auto-Submitted: auto-replied
4048 in accordance with these statements in the RFC:
4050 The auto-replied keyword:
4052 - SHOULD be used on messages sent in direct response to another
4053 message by an automatic process,
4055 - MUST NOT be used on manually-generated messages,
4057 - MAY be used on Delivery Status Notifications (DSNs) and Message
4058 Disposition Notifications (MDNs),
4060 - MUST NOT be used on messages generated by automatic or periodic
4061 processes, except for messages which are automatic responses to
4064 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4065 to the default Received: header definition.
4067 PH/12 Added log selector acl_warn_skipped (default on).
4069 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4070 variables because (a) they are in the wrong storage pool and (b) even if
4071 they were copied, it wouldn't work properly because of the caching.
4073 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4074 checking when decoding. Apparently there are clients that generate
4075 overlong encoded strings. Why am I not surprised?
4077 PH/15 If the first argument of "${if match_address" was not empty, but did not
4078 contain an "@" character, Exim crashed. Now it writes a panic log message
4079 and treats the condition as false.
4081 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4083 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4084 extension "envelope-auth". The code is finished and in agreement with
4085 other implementations, but there is no documentation so far and in fact,
4086 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4087 not changing the active code.
4089 Print executed "if" and "elsif" statements when debugging is used. This
4090 helps a great deal to understand what a filter does.
4092 Document more things not specified clearly in RFC3028. I had all this
4093 sorted out, when out of a sudden new issues came to my mind. Oops."
4095 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4098 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4099 canonical form (as documented). However, after a host name lookup from
4100 the IP address, check_host() was doing a simple string comparison with
4101 addresses acquired from the DNS when checking that the found name did
4102 have the original IP as one of its addresses. Since any found IPv6
4103 addresses are likely to be in abbreviated form, the comparison could
4104 fail. Luckily, there already exists a function for doing the comparison
4105 by converting both addresses to binary, so now that is used instead of
4106 the text comparison.
4108 PH/20 There was another similar case to PH/19, when a complete host name was
4109 given in a host list; looking up its IP address could give an abbreviated
4110 form, whereas the current host's name might or might not be abbreviated.
4111 The same fix has been applied.
4117 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4118 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4121 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4122 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4124 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4126 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4127 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4128 Also, for "Auto-Submitted", treat anything other than "no" as
4129 non-personal, in accordance with RFC 3834. (Previously it treated
4130 anything starting "auto-" as non-personal.)
4132 TF/01 The control=submission/name=... option had a problem with syntax
4133 errors if the name included a slash character. The /name= option
4134 now slurps the rest of the string, so it can include any characters
4135 but it must come last in the list of options (after /sender_retain
4138 PH/05 Some modifications to the interface to the fake nameserver for the new
4146 TK/01 Added the "success_on_redirect" address verification option. See
4147 NewStuff for rationale and an example.
4149 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4151 PH/02 Patch to exigrep to allow it to work on syslog lines.
4153 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4154 fread() to skip over the body file's header line, because in Cygwin the
4155 header line is locked and is inaccessible.
4157 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4158 co-exist for some time) to make it clear that it is the Exim ID that is
4159 referenced, not the Message-ID: header line.
4161 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4162 string_format() function, because snprintf() does not exist on all
4165 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4166 ${stat: expansion item.
4168 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4169 protocol synchronization error", to keep the pedants happy.
4171 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4172 well as for IRIX systems, when gcc is being used. See the host.c source
4175 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4177 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4180 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4181 transport and to the smtp transport in LMTP mode.
4183 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4185 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4186 run a filter in a subprocess. This could lead to confusion in subsequent
4187 lookups in the parent process. There should also be a search_tidyup() at
4188 the end of the subprocess.
4190 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4191 only if the host matched helo_try_verify_hosts, which caused the
4192 verification to occur when the EHLO/HELO command was issued. The ACL just
4193 tested the remembered result. Now, if a previous verification attempt has
4194 not happened, "verify = helo" does it there and then.
4196 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4198 TK/03 Fix log output including CR from clamd.
4200 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4201 exist provoked a memory error which could cause a segfault.
4203 PH/15 Installed PCRE 6.2
4205 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4207 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4208 of the problem. Specifically, suggested +O2 rather than +O1 for the
4211 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4213 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4214 record, the port was not correctly logged when the outgoing_port log
4215 selector was set (it logged the transort's default port).
4217 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4218 fallback_hosts, and "hosts" in the smtp transport.
4220 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4221 host errors such as "Connection refused".
4223 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4224 authentication with radiusclient 0.4.9:
4226 - Error returned from rc_read_config was caught wrongly
4227 - Username/password not passed on to radius server due to wrong length.
4229 The presumption is that some radiusclient API changes for 4.51/PH/17
4230 were not taken care of correctly. The code is still untested by me (my
4231 Linux distribution still has 0.3.2 of radiusclient), but it was
4232 contributed by a Radius user.
4234 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4235 expanding the "port" option of the smtp transport.
4237 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4238 while reading a MIME header. Thanks to Tom Hughes for a patch.
4240 PH/24 Include config.h inside local_scan.h so that configuration settings are
4243 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4244 that in an ACL for RCPT (for example), you can examine exactly what was
4247 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4248 commands, but it was not correctly comparing the address with the actual
4249 client host address. Thus, it would show the EHLO address in Received:
4250 header lines when this was not necessary.
4252 PH/27 Added the % operator to ${eval:}.
4254 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4255 it should be ignoring failures (because with -C, for example, it has lost
4256 privilege). It wasn't ignoring creation failures other than "already
4259 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4262 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4263 because an input line was too long, either on its own, or by virtue of
4264 too many continuations, the temporary file was not being removed, and the
4265 return code was incorrect.
4267 PH/31 Missing "BOOL" in function definition in filtertest.c.
4269 PH/32 Applied Sieve patches from the maintainer.
4271 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4273 PH/33 Added "verify = not_blind".
4275 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4276 Local/Makefile (with some defaults set). These are used in built scripts
4277 such as exicyclog, but they have never been used in the exim_install
4278 script (though there are many overriding facilities there). I have
4279 arranged that the exim_install script now takes note of these two
4282 PH/35 Installed configuration files for Dragonfly.
4284 PH/36 When a locally submitted message by a trusted user did not contain a
4285 From: header, and the sender address was obtained from -f or from an SMTP
4286 MAIL command, and the trusted user did not use -F to supply a sender
4287 name, $originator_name was incorrectly used when constructing a From:
4288 header. Furthermore, $originator_name was used for submission mode
4289 messages from external hosts without From: headers in a similar way,
4290 which is clearly wrong.
4292 PH/37 Added control=suppress_local_fixups.
4294 PH/38 When log_selector = +received_sender was set, and the addition of the
4295 sender made the log line's construction buffer exactly full, or one byte
4296 less than full, an overflow happened when the terminating "\n" was
4299 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4300 when the result of a list match is failure because a DNS lookup failed.
4302 PH/40 RM_COMMAND is now used in the building process.
4304 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4305 the "build-* directories that it finds.
4307 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4308 address in a domain literal was a prefix of an interface address.
4310 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4311 when verifying a sender address, unless rewrite_headers is false.
4313 PH/44 Wrote a long comment about why errors_to addresses are verified as
4314 recipients, not senders.
4316 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4317 the ratelimit ACL was added.
4319 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4321 PH/46 Added extra information about PostgreSQL errors to the error string.
4323 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4324 avoiding the need to install special zones in a real server. This is
4325 backwards compatible; if it can't find the fake resolver, it drops back.
4326 Thus, both old and new test suites can be run.
4328 TF/02 Added util/ratelimit.pl
4330 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4331 clock is set back in time.
4333 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4334 Candler <B.Candler@pobox.com>.
4336 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4337 OK for addresses that are the result of redirections.
4339 PH/48 A number of further additions for the benefit of the new test suite,
4340 including a fake gethostbyname() that interfaces to the fake DNS resolver
4343 TF/06 The fix for widen_domains has also been applied to qualify_single and
4344 search_parents which are the other dnslookup options that can cause
4347 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4350 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4352 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4353 service" if it encountered an MX record with an empty target hostname.
4354 The message is now "an MX or SRV record indicated no SMTP service".
4356 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4357 if the DNS of the sending site is misconfigured. This is quite a
4358 common situation. This change restores the behaviour of treating a
4359 helo verification defer as a failure.
4361 PH/51 If self=fail was set on a router, the bounce message did not include the
4362 actual error message.
4368 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4370 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4371 command itself ended in error, the underlying message about the transport
4372 filter timeout was being overwritten with the pipe command error. Now the
4373 underlying error message should be appended to the second error message.
4375 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4377 PH/02 Reduce the amount of output that "make" produces by default. Full output
4378 can still be requested.
4380 PH/03 The warning log line about a condition test deferring for a "warn" verb
4381 was being output only once per connection, rather than after each
4382 occurrence (because it was using the same function as for successful
4383 "warn" verbs). This seems wrong, so I have changed it.
4385 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4386 it should not have, which might have caused a crash in the right
4387 circumstances, but probably never did.
4389 PH/04 Installed a modified version of Tony Finch's patch to make submission
4390 mode fix the return path as well as the Sender: header line, and to
4391 add a /name= option so that you can make the user's friendly name appear
4394 TF/03 Added the control = fakedefer ACL modifier.
4396 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4397 Mark Lowes for thorough testing.
4399 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4401 TK/03 Merged latest SRS patch from Miles Wilton.
4403 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4404 with the definition in sysexits.h (which is #included earlier).
4405 Fortunately, Exim does not actually use EX_OK. The code used to try to
4406 preserve the sysexits.h value, by assuming that macro definitions were
4407 scanned for macro replacements. I have been disabused of this notion,
4408 so now the code just undefines EX_OK before #including unistd.h.
4410 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4411 in the smtp transport. When a block could not be written in a single
4412 write() function, the timeout was being re-applied to each part-write.
4413 This seems wrong - if the receiver was accepting one byte at a time it
4414 would take for ever. The timeout is now adjusted when this happens. It
4415 doesn't have to be particularly precise.
4417 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4418 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4420 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4421 without a domain if the check to <postmaster@domain> fails.
4423 SC/01 Eximstats: added -xls and the ability to specify output files
4424 (patch written by Frank Heydlauf).
4426 SC/02 Eximstats: use FileHandles for outputting results.
4428 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4430 SC/04 Eximstats: fixed display of large numbers with -nvr option
4432 SC/05 Eximstats: fixed merging of reports with empty tables.
4434 SC/06 Eximstats: added the -include_original_destination flag
4436 SC/07 Eximstats: removed tabs and trailing whitespace.
4438 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4440 TK/06 MBOX spool code: Add real "From " MBOX separator line
4441 so the .eml file is really in mbox format (even though
4442 most programs do not really care). Patch from Alex Miller.
4444 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4445 The latter is generated from $received_to and is only set if the
4446 message has one envelope recipient. SA can use these headers,
4447 obviously out-of-the-box. Patch from Alex Miller.
4449 PH/08 The ${def test on a variable was returning false if the variable's
4450 value was "0", contrary to what the specification has always said!
4451 The result should be true unless the variable is empty.
4453 PH/09 The syntax error of a character other than { following "${if
4454 def:variable_name" (after optional whitespace) was not being diagnosed.
4455 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4456 accidental colon was present, for example, could give incorrect results.
4458 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4459 result is used (not including appendfile, where other changes are about
4462 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4463 This involved changing a lot of size variables from int to off_t. It
4464 should work with maildirs and everything.
4466 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4467 spamd dying while we are connected to it.
4469 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4472 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4473 function for BDB 4.3.
4475 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4477 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4478 the sources for PCRE have been re-arranged and more files are now
4481 PH/15 The code I had for printing potentially long long variables in PH/11
4482 above was not the best (it lost precision). The length of off_t variables
4483 is now inspected at build time, and an appropriate printing format (%ld
4484 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4485 to be "long long int" or "long int". This is needed for the internal
4486 formatting function string_vformat().
4488 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4489 the configuration file to be ":syslog", then the script "guesses" where
4490 the logs files are, rather than using the compiled in default. In our
4491 case the guess is not the same as the compiled default, so the script
4492 suddenly stopped working when I started to use syslog. The patch checks
4493 to see if log_file_path is "". If so, it attempts to read it from exim
4494 with no configuration file to get the compiled in version, before it
4495 falls back to the previous guessing code."
4497 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4498 implementing BATV in an Exim configuration. See NewStuff for the gory
4501 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4502 Makefile that are specific to HP-UX.
4504 PH/18 If the "use_postmaster" option was set for a recipient callout together
4505 with the "random" option, the postmaster address was used as the MAIL
4506 FROM address for the random test, but not for the subsequent recipient
4507 test. It is now used for both.
4509 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4510 patch removes a few documentation additions to RFC 3028, because the
4511 latest draft now contains them. It adds the new en;ascii-case comparator
4512 and a new error check for 8bit text in MIME parts. Comparator and
4513 require names are now matched exactly. I enabled the subaddress
4514 extension, but it is not well tested yet (read: it works for me)."
4516 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4517 rework some of the code of TK/09 above to avoid the hardwired use of
4518 "%lld" and "long long". Replaced the call to snprintf() with a call to
4521 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4522 records point to non-existent hosts", "retry timeout exceeded", and
4523 "retry time not reached for any host after a long failure period".
4525 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4526 experimental DomainKeys support:
4528 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4529 (2) On an error such as an illegally used "control", the wrong name for
4530 the control was given.
4532 These problems did NOT occur unless DomainKeys support was compiled.
4534 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4536 PH/24 Added ${if match_ip condition.
4538 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4539 databases so that it will be absolutely obvious if a crash occurs in the
4540 DB library. This is a regular occurrence (often caused by mis-matched
4543 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4544 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4545 picked up on a user's system that detects such things. There doesn't seem
4546 to be a gcc warning option for this - only an attribute that has to be
4547 put on the function's prototype. It seems that in Fedora Core 4 they have
4548 set this on a number of new functions. No doubt there will be more in due
4551 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4552 specify a transport. However, if an address that was verified by such a
4553 router was the subject of a callout, Exim crashed because it tried to
4554 read the rcpt_include_affixes from the non-existent transport. Now it
4555 just assumes that the setting of that option is false. This bug was
4556 introduced by 4.51/PH/31.
4558 PH/28 Changed -d+all to exclude +memory, because that information is very
4559 rarely of interest, but it makes the output a lot bigger. People tend to
4560 do -d+all out of habit.
4562 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4563 code in os-type was giving problems when libc.so lives in lib64, like on
4566 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4567 aren't the modern standard, and it seems that some systems' include files
4568 don't always have them. Exim was already checking for some of the newer
4569 ones like T_AAAA, and defining it itself. I've added checks for all the
4570 record types that Exim uses.
4572 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4573 not automatically generating a new one, as it is supposed to. This
4574 prevented TLS from working. If the file did exist, but contained invalid
4575 data, a new version was generated, as expected. It was only the case of a
4576 non-existent file that was broken.
4578 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4579 with a change in libdomainkeys > 0.64.
4581 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4582 from DNS. If the selector record carries the flag, it now has
4583 precedence over the domain-wide flag.
4585 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4587 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4588 the use of a port name that isn't defined in /etc/services) occurred, the
4589 message was deferred as in a normal delivery, and thus remained on the
4590 spool, instead of being failed because of the mua_wrapper setting. This
4591 is now fixed, and I tidied up some of the mua_wrapper messages at the
4594 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4595 the messages in a hash of arrays rather than using individual hashes.
4596 This is a bit cleaner and results in dramatic memory savings, albeit
4597 at a slight CPU cost.
4599 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4600 as requested by Marc Sherman.
4602 SC/10 Eximstats: added histograms for user specified patterns as requested
4605 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4607 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4608 fopen() in the content-scanning modules that did not already have it.
4614 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4615 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4617 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4619 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4621 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4622 to test Sieve filters that use "vacation".
4624 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4625 that changes the way the GnuTLS parameters are stored in the cache file.
4626 The new format can be generated externally. For backward compatibility,
4627 if the data in the cache doesn't make sense, Exim assumes it has read an
4628 old-format file, and it generates new data and writes a new file. This
4629 means that you can't go back to an older release without removing the
4632 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4633 work if there are any delivery delays because "one_time" forces the
4634 parent to be marked "delivered", so its unseen clone is never tried
4635 again. For this reason, Exim now forbids the simultaneous setting of
4638 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4639 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4640 entry if you want to know the details. The fix, however, neglected to
4641 consider the case when local delivery batching is involved. The test for
4642 "previously delivered" was not happening when checking to see if an
4643 address could be batched with a previous (undelivered) one; under
4644 certain circumstances this could lead to multiple deliveries to the same
4647 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4648 in its include files, and this causes problems building Exim.
4650 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4651 header_syntax) but Exim was just ignoring anything given after a slash.
4652 In particular, this caused confusion with an attempt to use "verify =
4653 reverse_host_lookup/defer_ok". An error is now given when options are
4654 supplied for verify items that do not have them. (Maybe reverse_host_
4655 lookup should have a defer_ok option, but that's a different point.)
4657 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4658 defined by RFC 821) to 2048, because there were problems with some AUTH
4659 commands, and RFC 1869 says the size should be increased for extended
4660 SMTP commands that take arguments.
4662 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4665 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4666 "unknown" error; now it says that the functionality isn't in the binary.
4668 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4669 an address' error message when a string expansion fails (syntax or
4670 whatever). Otherwise the password may appear in the log. Following change
4671 PH/42 below, there is no longer a chance of it appearing in a bounce
4674 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4676 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4677 of its addresses was ever tried. (Bugzilla bug #2.)
4679 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4680 the result incorrectly in the debug output. (It correctly added a newline
4681 to what was transported.)
4683 TF/01 Added $received_time.
4685 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4686 commented out examples of how to interface to a virus scanner and to
4687 SpamAssassin. Also added commented examples of av_scanner and
4688 spamd_address settings.
4690 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4691 and controls are allowed in which ACLs. There were a couple of minor
4692 errors. Some of the entries in the conditions table (which is a table of
4693 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4694 negation of where the condition IS allowed.
4696 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4698 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4699 header file does not have a version number, so I've had to invent a new
4700 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4701 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4702 radiusclient), but it was contributed by a Radius user.
4704 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4705 files or directories, for OpenSSL.
4707 PH/19 When an Exim process that is running as root has to create an Exim log
4708 file, it does so in a subprocess that runs as exim:exim so as to get the
4709 ownership right at creation (otherwise, other Exim processes might see
4710 the file with the wrong ownership). There was no test for failure of this
4711 fork() call, which would lead to the process getting stuck as it waited
4712 for a non-existent subprocess. Forks do occasionally fail when resources
4713 run out. I reviewed all the other calls to fork(); they all seem to check
4716 PH/20 When checking for unexpected SMTP input at connect time (before writing
4717 the banner), Exim was not dealing correctly with a non-positive return
4718 from the read() function. If the client had disconnected by this time,
4719 the result was a log entry for a synchronization error with an empty
4720 string after "input=" when read() returned zero. If read() returned -1
4721 (an event I could not check), uninitialized data bytes were printed.
4722 There were reports of junk text (parts of files, etc) appearing after
4725 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4727 PH/22 Added support for macro redefinition, and (re)definition in between
4728 driver and ACL definitions.
4730 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4731 forgetting to use the resulting value; it was using the unexpanded value.
4733 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4734 hadn't been configured. The fix is from Juergen Kreileder, who
4735 understands it better than I do:
4737 "Here's what I see happening with three configured cyrus_sasl
4738 authenticators configured (plain, login, cram-md5):
4740 On startup auth_cyrus_sasl_init() gets called for each of these.
4741 This means three calls to sasl_listmech() without a specified mech_list.
4742 => SASL tests which mechs of all available mechs actually work
4743 => three warnings about OTP not working
4744 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4746 With the patch, sasl_listmech() also gets called three times. But now
4747 SASL's mech_list option is set to the server_mech specified in the the
4748 authenticator. Or in other words, the answer from sasl_listmech()
4749 gets limited to just the mech you're testing for (which is different
4751 => the return list contains just 'plain' or 'login', 'cram-md5' or
4752 nothing depending on the value of ob->server_mech.
4754 I've just tested the patch: Authentication still works fine,
4755 unavailable mechs specified in the exim configuration are still
4756 caught, and the auth.log warnings about OTP are gone."
4758 PH/25 When debugging is enabled, the contents of the command line are added
4759 to the debugging output, even when log_selector=+arguments is not
4762 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4763 answer is "GNU", and only if the return is "GNU/something" is the answer
4766 PH/27 $acl_verify_message is now set immediately after the failure of a
4767 verification in an ACL, and so is available in subsequent modifiers. In
4768 particular, the message can be preserved by coding like this:
4770 warn !verify = sender
4771 set acl_m0 = $acl_verify_message
4773 Previously, $acl_verify_message was set only while expanding "message"
4774 and "log_message" when a very denied access.
4776 PH/28 Modified OS/os.c-Linux with
4778 -#ifndef OS_LOAD_AVERAGE
4779 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4781 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4782 nomenclature these days.)
4784 PH/29 Installed patch from the Sieve maintainer that adds the options
4785 sieve_useraddress and sieve_subaddress to the redirect router.
4787 PH/30 In these circumstances:
4788 . Two addresses routed to the same list of hosts;
4789 . First host does not offer TLS;
4790 . First host accepts first address;
4791 . First host gives temporary error to second address;
4792 . Second host offers TLS and a TLS session is established;
4793 . Second host accepts second address.
4794 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4795 and peerdn, if requested) that were in fact used only for the second
4798 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4799 attention to any local part prefix or suffix that was matched by the
4800 router that accepted the address. It now behaves in the same way as it
4801 does for delivery: the affixes are removed from the local part unless
4802 rcpt_include_affixes is set on the transport.
4804 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4805 timeout during the DATA phase of an incoming message.
4807 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4808 applied a patch sanctioned by the Sieve maintainer.
4810 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4811 the uid or gid is negative. A case of a negative gid caused this to be
4812 noticed. The fix allows for either to be negative.
4814 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4815 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4818 PH/36 The change PH/12 above was broken. Fixed it.
4820 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4821 the grounds that routing the same address twice would always produce the
4822 same answer. This might have been true once, but it is certainly no
4823 longer true now. Routing a child address may depend on the previous
4824 routing that produced that child. Some complicated redirection strategies
4825 went wrong when messages had multiple recipients, and made Exim's
4826 behaviour dependent on the order in which the addresses were given.
4828 I have moved the duplicate checking until after the routing is complete.
4829 Exim scans the addresses that are assigned to local and remote
4830 transports, and removes any duplicates. This means that more work will be
4831 done, as duplicates will always all be routed, but duplicates are
4832 presumably rare, so I don't expect this is of any significance.
4834 For deliveries to pipes, files, and autoreplies, the duplicate checking
4835 still happens during the routing process, since they are not going to be
4838 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4839 It corrects a timeout issue with spamd. This is Ian's comment: "The
4840 background is that sometimes spamd either never reads data from a
4841 connection it has accepted, or it never writes response data. The exiscan
4842 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4843 blindly assumes that writes won't block so it may never time out."
4845 PH/39 Allow G after quota size as well as K and M.
4847 PH/40 The value set for $authenticated_id in an authenticator may not contain
4848 binary zeroes or newlines because the value is written to log lines and
4849 to spool files. There was no check on this. Now the value is run through
4850 the string_printing() function so that such characters are converted to
4851 printable escape sequences.
4853 PH/41 $message_linecount is a new variable that contains the total number of
4854 lines in the message. Compare $body_linecount, which is the count for the
4857 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4858 bounce and delay warning messages, except in certain special cases, which
4861 (a) An SMTP error message from a remote host;
4862 (b) A message specified in a :fail: redirection;
4863 (c) A message specified in a "fail" command in a system filter;
4864 (d) A message specified in a FAIL return from the queryprogram router;
4865 (e) A message specified by the cannot_route_message router option.
4867 In these cases only, Exim does include the error details in bounce and
4868 warning messages. There are also a few cases where bland messages such
4869 as "unrouteable address" or "local delivery error" are given.
4871 PH/43 $value is now also set for the "else" part of a ${run expansion.
4873 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4874 being worked on, but at least Exim now implements the latest version to
4877 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4878 process to complete was treated as a delivery failure, a timeout while
4879 writing the message to the pipe was logged, but erroneously treated as a
4880 successful delivery. Such timeouts include transport filter timeouts. For
4881 consistency with the overall process timeout, these timeouts are now
4882 treated as errors, giving rise to delivery failures by default. However,
4883 there is now a new Boolean option for the pipe transport called
4884 timeout_defer, which, if set TRUE, converts the failures into defers for
4885 both kinds of timeout. A transport filter timeout is now identified in
4888 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4889 systems where "make" and "gmake" are different, calling "gmake" at top
4890 level broke things. I've arranged for the value of $(MAKE) to be passed
4891 from the Makefile to this script so that it can call the same version of
4895 A note about Exim versions 4.44 and 4.50
4896 ----------------------------------------
4898 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4899 changes of various kinds. As a consequence, a big documentation update was
4900 needed. This delayed the release for rather longer than seemed good, especially
4901 in the light of a couple of (minor) security issues. Therefore, the changes
4902 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4903 release. So 4.44 and 4.50 are in effect two different branches that both start
4906 I have left the 4.50 change log unchanged; it contains all the changes since
4907 4.43. The change log for 4.44 is below; many of its items are identical to
4908 those for 4.50. This seems to be the most sensible way to preserve the
4909 historical information.
4915 1. Minor wording change to the doc/README.SIEVE file.
4917 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4918 computation of the current number of files was incorrect.
4920 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4921 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4924 4. Give more explanation in the error message when the command for a transport
4925 filter fails to execute.
4927 5. There are several places where Exim runs a non-Exim command in a
4928 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4929 was being done only for the command run by the queryprogram router. It is
4930 now done for all such subprocesses. The other cases are: ${run, transport
4931 filters, and the commands run by the lmtp and pipe transports.
4933 6. Added CONFIGURE_GROUP build-time option.
4935 7. Some older OS have a limit of 256 on the maximum number of file
4936 descriptors. Exim was using setrlimit() to set 1000 as a large value
4937 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4938 systems. I've change it so that if it can't get 1000, it tries for 256.
4940 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4941 was an oversight, and furthermore, ever since the addition of extra
4942 controls (e.g. 4.43/32), the checks on when to allow different forms of
4943 "control" were broken. There should now be diagnostics for all cases when a
4944 control that does not make sense is encountered.
4946 9. Added the /retain_sender option to "control=submission".
4948 10. $recipients is now available in the predata ACL (oversight).
4950 11. Tidy the search cache before the fork to do a delivery from a message
4951 received from the command line. Otherwise the child will trigger a lookup
4952 failure and thereby defer the delivery if it tries to use (for example) a
4953 cached ldap connection that the parent has called unbind on.
4955 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4956 of $address_data from the recipient verification was clobbered by the
4957 sender verification.
4959 13. The value of address_data from a sender verification is now available in
4960 $sender_address_data in subsequent conditions in the ACL statement.
4962 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4964 15. Added a new option "connect=<time>" to callout options, to set a different
4967 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4968 was its contents. (It was OK if the option was not defined at all.)
4970 17. A "Completed" log line is now written for messages that are removed from
4971 the spool by the -Mrm option.
4973 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4974 information about exactly what failed.
4976 19. Added -dd to debug only the daemon process.
4978 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4979 handles timeouts, both on the server side and network timeouts. Renamed the
4980 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4982 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4983 It is now set to "smtps".
4985 22. $host_address is now set to the target address during the checking of
4986 ignore_target_hosts.
4988 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4989 being passed; this would have caused $sender_host_name to have been used if
4990 matching the list had actually called for a host name (not very likely,
4991 since this list is usually IP addresses). A host name is now passed as
4994 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4995 code that specifies a non-restarting handler (typically sigaction() in
4996 modern systems) in an attempt to fix a rare and obscure crash bug.
4998 25. Narrowed the window for a race in the daemon that could cause it to ignore
4999 SIGCHLD signals. This is not a major problem, because they are used only to
5000 wake it up if nothing else does.
5002 26. A malformed maildirsize file could cause Exim to calculate negative values
5003 for the mailbox size or file count. Odd effects could occur as a result.
5004 The maildirsize information is now recalculated if the size or filecount
5007 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5008 support for a long time. Removed HAVE_SYS_VFS_H.
5010 28. Installed the latest version of exipick from John Jetmore.
5012 29. In an address list, if the pattern was not a regular expression, an empty
5013 subject address (from a bounce message) matched only if the pattern was an
5014 empty string. Non-empty patterns were not even tested. This was the wrong
5015 because it is perfectly reasonable to use an empty address as part of a
5016 database query. An empty address is now tested by patterns that are
5017 lookups. However, all the other forms of pattern expect the subject to
5018 contain a local part and a domain, and therefore, for them, an empty
5019 address still always fails if the pattern is not itself empty.
5021 30. Exim went into a mad DNS loop when attempting to do a callout where the
5022 host was specified on an smtp transport, and looking it up yielded more
5023 than one IP address.
5025 31. Re-factored the code for checking spool and log partition space into a
5026 function that finds that data and another that does the check. The former
5027 is then used to implement four new variables: $spool_space, $log_space,
5028 $spool_inodes, and $log_inodes.
5030 32. The RFC2047 encoding function was originally intended for short strings
5031 such as real names; it was not keeping to the 75-character limit for
5032 encoded words that the RFC imposes. It now respects the limit, and
5033 generates multiple encoded words if necessary. To be on the safe side, I
5034 have increased the buffer size for the ${rfc2047: expansion operator from
5037 33. It is now permitted to omit both strings after an "if" condition; if the
5038 condition is true, the result is "true". As before, when the second string
5039 is omitted, a false condition yields an empty string. This makes it less
5040 cumbersome to write custom ACL and router conditions.
5042 34. Failure to deliver a bounce message always caused it to be frozen, even if
5043 there was an errors_to setting on the router. The errors_to setting is now
5046 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5047 canonical form (fully expanded) before being placed in
5048 $sender_host_address.
5050 36. The table in the code that translates DNS record types into text (T_A to
5051 "A" for instance) was missing entries for NS and CNAME. It is just possible
5052 that this could have caused confusion if both these types were looked up
5053 for the same domain, because the text type is used as part of Exim's
5054 per-process caching. But the chance of anyone hitting this buglet seems
5057 37. The dnsdb lookup has been extended in a number of ways.
5059 (1) There is a new type, "zns", which walks up the domain tree until it
5060 finds some nameserver records. It should be used with care.
5062 (2) There is a new type, "mxh", which is like "mx" except that it returns
5063 just the host names, not the priorities.
5065 (3) It is now possible to give a list of domains (or IP addresses) to be
5066 looked up. The behaviour when one of the lookups defers can be
5067 controlled by a keyword.
5069 (4) It is now possible to specify the separator character for use when
5070 multiple records are returned.
5072 38. The dnslists ACL condition has been extended: it is now possible to supply
5073 a list of IP addresses and/or domains to be looked up in a particular DNS
5076 39. Added log_selector=+queue_time_overall.
5078 40. When running the queue in the test harness, wait just a tad after forking a
5079 delivery process, to get repeatability of debugging output.
5081 41. Include certificate and key file names in error message when GnuTLS fails
5082 to set them up, because the GnuTLS error message doesn't include the name
5083 of the failing file when there is a problem reading it.
5085 42. Allow both -bf and -bF in the same test run.
5087 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5089 44. The "Exiscan patch" is now merged into the mainline Exim source.
5091 45. Sometimes the final signoff response after QUIT could fail to get
5092 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5093 < 0 before doing a fflush(). This bug looks as though it goes back to the
5094 introduction of TLS in release 3.20, but "sometimes" must have been rare
5095 because the tests only now provoked it.
5097 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5098 (this can affect the format of dates).
5100 47. exim_tidydb, when checking for the continued existence of a message for
5101 which it has found a message-specific retry record, was not finding
5102 messages that were in split spool directories. Consequently, it was
5103 deleting retry records that should have stayed in existence.
5105 48. Steve fixed some bugs in eximstats.
5107 49. The SPA authentication driver was not abandoning authentication and moving
5108 on to the next authenticator when an expansion was forced to fail,
5109 contradicting the general specification for all authenticators. Instead it
5110 was generating a temporary error. It now behaves as specified.
5112 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5113 (the order specifies the preference for clients). The order is now AES256,
5114 AES128, 3DES, ARCFOUR128.
5116 51. Small patch to Sieve code - explicitly set From: when generating an
5119 52. Exim crashed if a remote delivery caused a very long error message to be
5120 recorded - for instance if somebody sent an entire SpamAssassin report back
5121 as a large number of 550 error lines. This bug was coincidentally fixed by
5122 increasing the size of one of Exim's internal buffers (big_buffer) that
5123 happened as part of the Exiscan merge. However, to be on the safe side, I
5124 have made the code more robust (and fixed the comments that describe what
5127 53. Now that there can be additional text after "Completed" in log lines (if
5128 the queue_time_overall log selector is set), a one-byte patch to exigrep
5129 was needed to allow it to recognize "Completed" as not the last thing in
5132 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5133 patch that reportedly fixes this has been added. I am not expert enough to
5134 create a test for it. This is what the patch creator wrote:
5136 "I found a little strange behaviour of ldap code when working with
5137 Windows 2003 AD Domain, where users was placed in more than one
5138 Organization Units. When I tried to give exim partial DN, the exit code
5139 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5140 But simultaneously result of request was absolutely normal ldap result,
5141 so I produce this patch..."
5143 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5144 REFERENCE, so I have modified the code to exclude the patch when that macro
5147 55. Some experimental protocols are using DNS PTR records for new purposes. The
5148 keys for these records are domain names, not reversed IP addresses. The
5149 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5150 leaves it alone. Component reversal etc. now happens only for IP addresses.
5153 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5155 57. Double the size of the debug message buffer (to 2048) so that more of very
5156 long debug lines gets shown.
5158 58. The exicyclog utility now does better if the number of log files to keep
5159 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5161 59. Two changes related to the smtp_active_hostname option:
5163 (1) $smtp_active_hostname is now available as a variable.
5164 (2) The default for smtp_banner uses $smtp_active_hostname instead
5165 of $primary_hostname.
5167 60. The host_aton() function is supposed to be passed a string that is known
5168 to be a valid IP address. However, in the case of IPv6 addresses, it was
5169 not checking this. This is a hostage to fortune. Exim now panics and dies
5170 if the condition is not met. A case was found where this could be provoked
5171 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5172 components; fortuitously, this particular loophole had already been fixed
5173 by change 4.50/55 above.
5175 If there are any other similar loopholes, the new check in host_aton()
5176 itself should stop them being exploited. The report I received stated that
5177 data on the command line could provoke the exploit when Exim was running as
5178 exim, but did not say which command line option was involved. All I could
5179 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5180 running as the user.
5183 61. There was a buffer overflow vulnerability in the SPA authentication code
5184 (which came originally from the Samba project). I have added a test to the
5185 spa_base64_to_bits() function which I hope fixes it.
5188 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5189 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5191 63. The daemon start-up calls getloadavg() while still root for those OS that
5192 need the first call to be done as root, but it missed one case: when
5193 deliver_queue_load_max is set with deliver_drop_privilege. This is
5194 necessary for the benefit of the queue runner, because there is no re-exec
5195 when deliver_drop_privilege is set.
5197 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5198 This has been fixed.
5200 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5201 was in use, was not putting the data itself into the right store pool;
5202 consequently, it could be overwritten for a subsequent message in the same
5203 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5206 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5208 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5209 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5210 the function were treating the return as a boolean value, which happened to
5211 work because 0=false and not-0=true, but is not correct code.
5213 68. The host_aton() function was not handling scoped IPv6 addresses (those
5214 with, for example, "%eth0" on the end) correctly.
5216 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5217 negated items (that is, ~something) in unsigned ints. Some compilers
5218 apparently mutter when there is no cast.
5220 70. If an address verification called from an ACL failed, and did not produce a
5221 user-specific message (i.e. there was only a "system" message), nothing was
5222 put in $acl_verify_message. In this situation, it now puts the system
5225 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5226 session; change 4.31/43 added the unwanted input to the log line - except
5227 that it did not do this in the start of session case. It now does.
5229 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5230 This is wrong and can cause the other end to generate a synchronization
5231 error if it is another Exim or anything else that does the synchronization
5232 check. A QUIT command is no longer sent after a timeout.
5234 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5235 during host lookups.
5237 74. The defer_ok option of callout verification was not working if it was used
5238 when verifying addresses in header lines, that is, for this case:
5240 verify = header_sender/callout=defer_ok
5242 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5243 those file descriptors could be used for SMTP connections. If anything
5244 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5245 could be sent to the SMTP client, causing chaos. The daemon now opens
5246 stdin, stdout, and stderr to /dev/null when it puts itself into the
5249 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5250 log by default. The user can override this with suitable Perl magic.
5252 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5253 the log message when discard triggers, was not working for the DATA ACL or
5254 for the non-SMTP ACL.
5256 78. Error message wording change in sieve.c.
5258 79. If smtp_accept_max_per_host was set, the number of connections could be
5259 restricted to fewer than expected, because the daemon was trying to set up
5260 a new connection before checking whether the processes handling previous
5261 connections had finished. The check for completed processes is now done
5262 earlier. On busy systems, this bug wouldn't be noticed because something
5263 else would have woken the daemon, and it would have reaped the completed
5266 80. If a message was submitted locally by a user whose login name contained one
5267 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5268 It caused a spool format error. I have fixed the spool reading code. A
5269 related problem was that the "from" clause in the Received: line became
5270 illegal because of the space(s). It is now covered by ${quote_local_part.
5272 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5275 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5277 83. Patch from Sieve maintainer to fix unterminated string problem in
5278 "vacation" handling.
5280 84. Some minor changes to the Linux configuration files to help with other
5281 OS variants using glibc.
5283 85. One more patch for Sieve to update vacation handling to latest spec.
5286 ----------------------------------------------------
5287 See the note above about the 4.44 and 4.50 releases.
5288 ----------------------------------------------------
5294 1. Change 4.43/35 introduced a bug that caused file counts to be
5295 incorrectly computed when quota_filecount was set in an appendfile
5298 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5299 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5302 3. Give more explanation in the error message when the command for a transport
5303 filter fails to execute.
5305 4. There are several places where Exim runs a non-Exim command in a
5306 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5307 was being done only for the command run by the queryprogram router. It is
5308 now done for all such subprocesses. The other cases are: ${run, transport
5309 filters, and the commands run by the lmtp and pipe transports.
5311 5. Some older OS have a limit of 256 on the maximum number of file
5312 descriptors. Exim was using setrlimit() to set 1000 as a large value
5313 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5314 systems. I've change it so that if it can't get 1000, it tries for 256.
5316 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5317 was an oversight, and furthermore, ever since the addition of extra
5318 controls (e.g. 4.43/32), the checks on when to allow different forms of
5319 "control" were broken. There should now be diagnostics for all cases when a
5320 control that does not make sense is encountered.
5322 7. $recipients is now available in the predata ACL (oversight).
5324 8. Tidy the search cache before the fork to do a delivery from a message
5325 received from the command line. Otherwise the child will trigger a lookup
5326 failure and thereby defer the delivery if it tries to use (for example) a
5327 cached ldap connection that the parent has called unbind on.
5329 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5330 of $address_data from the recipient verification was clobbered by the
5331 sender verification.
5333 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5334 was its contents. (It was OK if the option was not defined at all.)
5336 11. A "Completed" log line is now written for messages that are removed from
5337 the spool by the -Mrm option.
5339 12. $host_address is now set to the target address during the checking of
5340 ignore_target_hosts.
5342 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5343 being passed; this would have caused $sender_host_name to have been used if
5344 matching the list had actually called for a host name (not very likely,
5345 since this list is usually IP addresses). A host name is now passed as
5348 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5349 code that specifies a non-restarting handler (typically sigaction() in
5350 modern systems) in an attempt to fix a rare and obscure crash bug.
5352 15. Narrowed the window for a race in the daemon that could cause it to ignore
5353 SIGCHLD signals. This is not a major problem, because they are used only to
5354 wake it up if nothing else does.
5356 16. A malformed maildirsize file could cause Exim to calculate negative values
5357 for the mailbox size or file count. Odd effects could occur as a result.
5358 The maildirsize information is now recalculated if the size or filecount
5361 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5362 support for a long time. Removed HAVE_SYS_VFS_H.
5364 18. Updated exipick to current release from John Jetmore.
5366 19. Allow an empty sender to be matched against a lookup in an address list.
5367 Previously the only cases considered were a regular expression, or an
5370 20. Exim went into a mad DNS lookup loop when doing a callout where the
5371 host was specified on the transport, if the DNS lookup yielded more than
5374 21. The RFC2047 encoding function was originally intended for short strings
5375 such as real names; it was not keeping to the 75-character limit for
5376 encoded words that the RFC imposes. It now respects the limit, and
5377 generates multiple encoded words if necessary. To be on the safe side, I
5378 have increased the buffer size for the ${rfc2047: expansion operator from
5381 22. Failure to deliver a bounce message always caused it to be frozen, even if
5382 there was an errors_to setting on the router. The errors_to setting is now
5385 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5386 canonical form (fully expanded) before being placed in
5387 $sender_host_address.
5389 24. Updated eximstats to version 1.33
5391 25. Include certificate and key file names in error message when GnuTLS fails
5392 to set them up, because the GnuTLS error message doesn't include the name
5393 of the failing file when there is a problem reading it.
5395 26. Expand error message when OpenSSL has problems setting up cert/key files.
5398 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5399 (this can affect the format of dates).
5401 28. exim_tidydb, when checking for the continued existence of a message for
5402 which it has found a message-specific retry record, was not finding
5403 messages that were in split spool directories. Consequently, it was
5404 deleting retry records that should have stayed in existence.
5406 29. eximstats updated to version 1.35
5407 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5408 1.35 - bugfix such that pie charts by volume are generated correctly
5410 30. The SPA authentication driver was not abandoning authentication and moving
5411 on to the next authenticator when an expansion was forced to fail,
5412 contradicting the general specification for all authenticators. Instead it
5413 was generating a temporary error. It now behaves as specified.
5415 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5416 (the order specifies the preference for clients). The order is now AES256,
5417 AES128, 3DES, ARCFOUR128.
5419 31. Small patch to Sieve code - explicitly set From: when generating an
5422 32. Exim crashed if a remote delivery caused a very long error message to be
5423 recorded - for instance if somebody sent an entire SpamAssassin report back
5424 as a large number of 550 error lines. This bug was coincidentally fixed by
5425 increasing the size of one of Exim's internal buffers (big_buffer) that
5426 happened as part of the Exiscan merge. However, to be on the safe side, I
5427 have made the code more robust (and fixed the comments that describe what
5430 33. Some experimental protocols are using DNS PTR records for new purposes. The
5431 keys for these records are domain names, not reversed IP addresses. The
5432 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5433 leaves it alone. Component reversal etc. now happens only for IP addresses.
5436 34. The host_aton() function is supposed to be passed a string that is known
5437 to be a valid IP address. However, in the case of IPv6 addresses, it was
5438 not checking this. This is a hostage to fortune. Exim now panics and dies
5439 if the condition is not met. A case was found where this could be provoked
5440 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5441 components; fortuitously, this particular loophole had already been fixed
5442 by change 4.50/55 or 4.44/33 above.
5444 If there are any other similar loopholes, the new check in host_aton()
5445 itself should stop them being exploited. The report I received stated that
5446 data on the command line could provoke the exploit when Exim was running as
5447 exim, but did not say which command line option was involved. All I could
5448 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5449 running as the user.
5452 35. There was a buffer overflow vulnerability in the SPA authentication code
5453 (which came originally from the Samba project). I have added a test to the
5454 spa_base64_to_bits() function which I hope fixes it.
5457 36. The daemon start-up calls getloadavg() while still root for those OS that
5458 need the first call to be done as root, but it missed one case: when
5459 deliver_queue_load_max is set with deliver_drop_privilege. This is
5460 necessary for the benefit of the queue runner, because there is no re-exec
5461 when deliver_drop_privilege is set.
5463 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5464 was in use, was not putting the data itself into the right store pool;
5465 consequently, it could be overwritten for a subsequent message in the same
5466 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5469 38. Sometimes the final signoff response after QUIT could fail to get
5470 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5471 < 0 before doing a fflush(). This bug looks as though it goes back to the
5472 introduction of TLS in release 3.20, but "sometimes" must have been rare
5473 because the tests only now provoked it.
5479 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5480 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5481 Now it returns an integer. A number of calls were still expecting a T/F
5482 return. Fortuitously, in all cases, the tests worked in OK situations,
5483 which is the norm. However, things would have gone wrong on any write
5484 failures on the smtp file descriptor. This function is used when sending
5485 messages over SMTP and also when doing verify callouts.
5487 2. When Exim is called to do synchronous delivery of a locally submitted
5488 message (the -odf or -odi options), it no longer closes stderr before doing
5491 3. Implemented the mua_wrapper option.
5493 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5495 5. Implemented the functions header_remove(), header_testname(),
5496 header_add_at_position(), and receive_remove_recipient(), and exported them
5499 6. If an ACL "warn" statement specified the addition of headers, Exim already
5500 inserted X-ACL-Warn: at the start if there was no header name. However, it
5501 was not making this test for the second and subsequent header lines if
5502 there were newlines in the string. This meant that an invalid header could
5503 be inserted if Exim was badly configured.
5505 7. Allow an ACL "warn" statement to add header lines at the start or after all
5506 the Received: headers, as well as at the end.
5508 8. Added the rcpt_4xx retry error code.
5510 9. Added postmaster_mailfrom=xxx to callout verification option.
5512 10. Added mailfrom=xxxx to the callout verification option, for verify=
5515 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5516 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5518 12. Inserted some casts to stop certain compilers warning when using pointer
5519 differences as field lengths or precisions in printf-type calls (mostly
5520 affecting debugging statements).
5522 13. Added optional readline() support for -be (dynamically loaded).
5524 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5525 same clock tick as a message's arrival, so that its received time was the
5526 same as the "first fail" time on the retry record, and that message
5527 remained on the queue past the ultimate address timeout, every queue runner
5528 would try a delivery (because it was past the ultimate address timeout) but
5529 after another failure, the ultimate address timeout, which should have then
5530 bounced the address, did not kick in. This was a "< instead of <=" error;
5531 in most cases the first failure would have been in the next clock tick
5532 after the received time, and all would be well.
5534 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5535 being recognized when the domain list was tested by the match_domain
5536 condition in an expansion string.
5538 16. Added the ${str2b64: operator.
5540 17. Exim was always calling setrlimit() to set a large limit for the number of
5541 processes, without checking whether the existing limit was already
5542 adequate. (It did check for the limit on file descriptors.) Furthermore,
5543 errors from getrlimit() and setrlimit() were being ignored. Now they are
5544 logged to the main and panic logs, but Exim does carry on, to try to do its
5545 job under whatever limits there are.
5547 18. Imported PCRE 5.0.
5549 19. Trivial typo in log message " temporarily refused connection" (the leading
5552 20. If the log selector return_path_on_delivery was set and an address was
5553 redirected to /dev/null, the delivery process crashed because it assumed
5554 that a return path would always be set for a "successful" delivery. In this
5555 case, the whole delivery is bypassed as an optimization, and therefore no
5558 21. Internal re-arrangement: the function for sending a challenge and reading
5559 a response while authentication was assuming a zero-terminated challenge
5560 string. It's now changed to take a pointer and a length, to allow for
5561 binary data in such strings.
5563 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5565 23. Exim was not respecting finduser_retries when seeking the login of the
5566 uid under which it was called; it was always trying 10 times. (The default
5567 setting of finduser_retries is zero.) Also, it was sleeping after the final
5568 failure, which is pointless.
5570 24. Implemented tls_on_connect_ports.
5572 25. Implemented acl_smtp_predata.
5574 26. If the domain in control=submission is set empty, Exim assumes that the
5575 authenticated id is a complete email address when it generates From: or
5576 Sender: header lines.
5578 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5579 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5580 chown and chgrp in /bin and hostname in /usr/bin.
5582 28. Exim was keeping the "process log" file open after each use, just as it
5583 does for the main log. This opens the possibility of it remaining open for
5584 long periods when the USR1 signal hits a daemon. Occasional processlog
5585 errors were reported, that could have been caused by this. Anyway, it seems
5586 much more sensible not to leave this file open at all, so that is what now
5589 29. The long-running daemon process does not normally write to the log once it
5590 has entered its main loop, and it closes the log before doing so. This is
5591 so that log files can straightforwardly be renamed and moved. However,
5592 there are a couple of unusual error situations where the daemon does write
5593 log entries, and I had neglected to close the log afterwards.
5595 30. The text of an SMTP error response that was received during a remote
5596 delivery was being truncated at 512 bytes. This is too short for some of
5597 the long messages that one sometimes sees. I've increased the limit to
5600 31. It is now possible to make retry rules that apply only when a message has a
5601 specific sender, in particular, an empty sender.
5603 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5604 it possible to be selective about when SMTP synchronization is enforced.
5606 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5608 32. Implemented hosts_connection_nolog.
5610 33. Added an ACL for QUIT.
5612 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5615 35. Added mailbox_size and mailbox_filecount to appendfile.
5617 36. Added control = no_multiline_responses to ACLs.
5619 37. There was a bug in the logic of the code that waits for the clock to tick
5620 in the case where the clock went backwards by a substantial amount such
5621 that the microsecond fraction of "now" was more than the microsecond
5622 fraction of "then" (but the whole seconds number was less).
5624 38. Added support for the libradius Radius client library this is found on
5625 FreeBSD (previously only the radiusclient library was supported).
5631 1. When certain lookups returned multiple values in the form name=value, the
5632 quoting of the values was not always being done properly. Specifically:
5633 (a) If the value started with a double quote, but contained no whitespace,
5635 (b) If the value contained whitespace other than a space character (i.e.
5636 tabs or newlines or carriage returns) it was not quoted.
5637 This fix has been applied to the mysql and pgsql lookups by writing a
5638 separate quoting function and calling it from the lookup code. The fix
5639 should probably also be applied to nisplus, ibase and oracle lookups, but
5640 since I cannot test any of those, I have not disturbed their existing code.
5642 2. A hit in the callout cache for a specific address caused a log line with no
5643 reason for rejecting RCPT. Now it says "Previous (cached) callout
5644 verification failure".
5646 3. There was an off-by-one bug in the queryprogram router. An over-long
5647 return line was truncated at 256 instead of 255 characters, thereby
5648 overflowing its buffer with the terminating zero. As well as fixing this, I
5649 have increased the buffer size to 1024 (and made a note to document this).
5651 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5652 when Exim is waiting for an SMTP response from a remote server, Exim
5653 restarts its select() call on the socket, thereby resetting its timeout.
5654 This is not a problem when such interrupts are rare. Somebody set up a cron
5655 job to run exiwhat every 2 minutes, which is less than the normal select()
5656 timeout (5 or 10 minutes). This meant that the select() timeout never
5657 kicked in because it was always reset. I have fixed this by comparing the
5658 time when an interrupt arrives with the time at the start of the first call
5659 to select(). If more time than the timeout has elapsed, the interrupt is
5660 treated as a timeout.
5662 5. Some internal re-factoring in preparation for the addition of Sieve
5663 extensions (by MH). In particular, the "personal" test is moved to a
5664 separate function, and given an option for scanning Cc: and Bcc: (which is
5665 not set for Exim filters).
5667 6. When Exim created an email address using the login of the caller as the
5668 local part (e.g. when creating a From: or Sender: header line), it was not
5669 quoting the local part when it contained special characters such as @.
5671 7. Installed new OpenBSD configuration files.
5673 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5674 try to make them clearer.
5676 9. Callout options, other than the timeout value, were being ignored when
5677 verifying sender addresses in header lines. For example, when using
5679 verify = header_sender/callout=no_cache
5681 the cache was (incorrectly) being used.
5683 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5684 only the Cygwin environment.
5686 11. When return_path_on_delivery was set as a log selector, if different remote
5687 addresses in the same message used different return paths and parallel
5688 remote delivery occurred, the wrong values would sometimes be logged.
5689 (Whenever a remote delivery process finished, the return path value from
5690 the most recently started remote delivery process was logged.)
5692 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5693 lines when AUTH and/or TLS are in use. This is the "received protocol"
5694 field. Exim used to use "asmtp" for authenticated SMTP, without any
5695 indication (in the protocol name) for TLS use. Now it follows the RFC and
5696 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5697 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5698 names appear in log lines as well as in Received: header lines.
5700 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5701 extensions, and comparison tests, and to fix some bugs.
5703 14. Changes to the "personal" filter test:
5705 (1) The test was buggy in that it was just doing the equivalent of
5706 "contains" tests on header lines. For example, if a user's address was
5707 anne@some.where, the "personal" test would incorrectly be true for
5709 To: susanne@some.where
5711 This test is now done by extracting each address from the header in turn,
5712 and checking the entire address. Other tests that are part of "personal"
5713 are now done using regular expressions (for example, to check local parts
5714 of addresses in From: header lines).
5716 (2) The list of non-personal local parts in From: addresses has been
5717 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5718 taken from the Sieve specification recommendations.
5720 (3) If the message contains any header line starting with "List-" it is
5721 treated as non-personal.
5723 (4) The test for "circular" in the Subject: header line has been removed
5724 because it now seems ill-conceived.
5726 15. Minor typos in src/EDITME comments corrected.
5728 16. Installed latest exipick from John Jetmore.
5730 17. If headers_add on a router specified a text string that was too long for
5731 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5732 of string_sprintf() is now avoided.
5734 18. $message_body_size was not set (it was always zero) when running the DATA
5735 ACL and the local_scan() function.
5737 19. For the "mail" command in an Exim filter, no default was being set for
5738 the once_repeat time, causing a random time value to be used if "once" was
5739 specified. (If the value happened to be <= 0, no repeat happened.) The
5740 default is now 0s, meaning "never repeat". The "vacation" command was OK
5741 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5742 (I found it when inspecting the code).
5744 20. There is now an overall timeout for performing a callout verification. It
5745 defaults to 4 times the callout timeout, which applies to individual SMTP
5746 commands during the callout. The overall timeout applies when there is more
5747 than one host that can be tried. The timeout is checked before trying the
5748 next host. This prevents very long delays if there are a large number of
5749 hosts and all are timing out (e.g. when the network connections are timing
5750 out). The value of the overall timeout can be changed by specifying an
5751 additional sub-option for "callout", called "maxwait". For example:
5753 verify = sender/callout=5s,maxwait=20s
5755 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5756 to the end before writing, but this should make it even safer).
5758 22. Exim was forgetting that it had advertised PIPELINING for the second and
5759 subsequent messages on an SMTP connection. It was also not resetting its
5760 memory on STARTTLS and an internal HELO.
5762 23. When Exim logs an SMTP synchronization error within a session, it now
5763 records whether PIPELINING has been advertised or not.
5765 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5766 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5767 rather than long int.
5769 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5771 26. Added the never_mail option to autoreply.
5777 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5778 crash if the getsockname() call failed; this can happen if a connection is
5779 closed very soon after it is established. The problem was simply in the
5780 order in which certain operations were done, causing Exim to try to write
5781 to the SMTP stream before it had set up the file descriptor. The bug has
5782 been fixed by making things happen in the correct order.
5788 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5789 before the connection was closed, thus losing the rejection response.
5791 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5792 some early Solaris releases, but causes trouble in current releases where
5793 socklen_t is defined.
5795 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5798 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5801 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5802 that releasing the top of it at the end releases what was used for sub-
5803 expansions (unless the block got too big). However, discard this block if
5804 the first thing is a variable or header, so that we can use its block when
5805 it is dynamic (useful for very large $message_headers, for example).
5807 6. Lookups now cache *every* query, not just the most recent. A new, separate
5808 store pool is used for this. It can be recovered when all lookup caches are
5809 flushed. Lookups now release memory at the end of their result strings.
5810 This has involved some general refactoring of the lookup sources.
5812 7. Some code has been added to the store_xxx() functions to reduce the amount
5813 of flapping under certain conditions.
5815 8. log_incoming_interface used to affect only the <= reception log lines. Now
5816 it causes the local interface and port to be added to several more SMTP log
5817 lines, for example "SMTP connection from", and rejection lines.
5819 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5821 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5823 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5824 could be overwritten at the end of the current message (or the start of a
5825 new message if it was set in a HELO ACL). The value is now preserved for
5826 the duration of the SMTP connection.
5828 12. If a transport had a headers_rewrite setting, and a matching header line
5829 contained an unqualified address, that address was qualified, even if it
5830 did not match any rewriting rules. The underlying bug was that the values
5831 of the flags that permit the existence of unqualified sender and recipient
5832 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5833 non-local messages, and by -bnq for local messages) were not being
5834 preserved with the message after it was received.
5836 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5837 "next input=" as part of the text comprising the host identity instead of
5838 the correct text. The code was using the same buffer for two different
5839 strings. However, depending on which order the printing function evaluated
5840 its arguments, the bug did not always show up. Under Linux, for example, my
5841 test suite worked just fine.
5843 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5844 doesn't work with some older versions of Perl. It has been changed to "my",
5845 which in any case is probably the better facility to use.
5847 15. A really picky compiler found some instances of statements for creating
5848 error messages that either had too many or two few arguments for the format
5851 16. The size of the buffer for calls to the DNS resolver has been increased
5852 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5853 for addresses that have a lot of PTR records. This alleviates a problem; it
5854 does not fully solve it.
5856 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5857 buffer now truncates the list and logs the incident, which is the same
5858 action as happens when Exim is looking up a host name and its aliases.
5859 Previously in this situation something unpredictable would happen;
5860 sometimes it was "internal error: store_reset failed".
5862 18. If a server dropped the connection unexpectedly when an Exim client was
5863 using GnuTLS and trying to read a response, the client delivery process
5864 crashed while trying to generate an error log message.
5866 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5867 string, for example:
5869 warn message = H1: something\nH2: something
5871 the text was added as a single header line from Exim's point of view
5872 though it ended up OK in the delivered message. However, searching for the
5873 second and subsequent header lines using $h_h2: did not work. This has been
5874 fixed. Similarly, if a system filter added multiple headers in this way,
5875 the routers could not see them.
5877 20. Expanded the error message when iplsearch is called with an invalid key to
5878 suggest using net-iplsearch in a host list.
5880 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5881 ACLs are no longer actually imposed (and a message to that effect is
5884 22. If a "gecos" field in a passwd entry contained escaped characters, in
5885 particular, if it contained a \" sequence, Exim got it wrong when building
5886 a From: or a Sender: header from that name. A second bug also caused
5887 incorrect handling when an unquoted " was present following a character
5888 that needed quoting.
5890 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5891 was not being matched caselessly.
5893 24. Arranged for all hyphens in the exim.8 source to be escaped with
5896 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5897 properly. Recipient callout cache records were still being keyed to include
5898 the sender, even when use_sender was set false. This led to far more
5899 callouts that were necessary. The sender is no longer included in the key
5900 when use_sender is false.
5902 26. Added "control = submission" modifier to ACLs.
5904 27. Added the ${base62d: operator to decode base 62 numbers.
5906 28. dnsdb lookups can now access SRV records.
5908 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5909 the configuration file.
5911 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5912 (-v) mode. This makes the output for a verbose queue run more intelligible.
5914 31. Added a use_postmaster feature to recipient callouts.
5916 32. Added the $body_zerocount variable, containing the number of binary zero
5917 bytes in the message body.
5919 33. The time of last modification of the "new" subdirectory is now used as the
5920 "mailbox time last read" when there is a quota error for a maildir
5923 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5925 35. Added +ignore_unknown as a special item in host lists.
5927 36. Code for decoding IPv6 addresses in host lists is now included, even if
5928 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5929 address was recognized as an IP address, but was then not correctly decoded
5930 into binary, causing unexpected and incorrect effects when compared with
5937 1. Very minor rewording of debugging text in manualroute to say "list of
5938 hosts" instead of "hostlist".
5940 2. If verify=header_syntax was set, and a header line with an unqualified
5941 address (no domain) and a large number of spaces between the end of the
5942 name and the colon was received, the reception process suffered a buffer
5943 overflow, and (when I tested it) crashed. This was caused by some obsolete
5944 code that should have been removed. The fix is to remove it!
5946 3. When running in the test harness, delay a bit after writing a bounce
5947 message to get a bit more predictability in the log output.
5949 4. Added a call to search_tidyup() just before forking a reception process. In
5950 theory, someone could use a lookup in the expansion of smtp_accept_max_
5951 per_host which, without the tidyup, could leave open a database connection.
5953 5. Added the variables $recipient_data and $sender_data which get set from a
5954 lookup success in an ACL "recipients" or "senders" condition, or a router
5955 "senders" option, similar to $domain_data and $local_part_data.
5957 6. Moved the writing of debug_print from before to after the "senders" test
5960 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5961 problems for message scanning, either using a data ACL, or using
5962 local_scan() because the Received: header was not generated till after they
5963 were called (in order to set the time as the time of reception completion).
5964 I have revised the way this works. The header is now generated after the
5965 body is received, but before the ACL or local_scan() are called. After they
5966 are run, the timestamp in the header is updated.
5972 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5973 before starting a queue runner without re-exec. This happened only when
5974 deliver_drop_privilege was set or when the Exim user was set to root. The
5975 effect of the bug was that timeouts during subsequent deliveries caused
5976 crashes instead of being properly handled. The handler is now left at its
5977 default (and expected) setting.
5979 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5980 message, again when deliver_drop_privilege is set or Exim is run as root.
5981 The bug described in (1) was not present in this case, but the tidying up
5982 of the other signals was missing. I have made the two cases consistent.
5984 3. The ignore_target_hosts setting on a manualroute router was being ignored
5985 for hosts that were looked up using the /MX notation.
5987 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5990 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5991 operated on the sender address. After changing the $sender_address to <>
5992 for the sender address verify, Exim was re-instated it as the original
5993 (before rewriting) address, but remembering that it had rewritten it, so it
5994 wasn't rewriting it again. This bug also had the effect of breaking the
5995 sender address verification caching when the sender address was rewritten.
5997 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5998 This has been changed so that if the ip literal address matches
5999 ignore_target_hosts, the router declines.
6001 7. Added expansion conditions match_domain, match_address, and match_local_
6002 part (NOT match_host).
6004 8. The placeholder for the Received: header didn't have a length field set.
6006 9. Added code to Exim itself and to exim_lock to test for a specific race
6007 condition that could lead to file corruption when using MBX delivery. The
6008 issue is with the lockfile that is created in /tmp. If this file is removed
6009 after a process has opened it but before that process has acquired a lock,
6010 there is the potential for a second process to recreate the file and also
6011 acquire a lock. This could lead to two Exim processes writing to the file
6012 at the same time. The added code performs the same test as UW imapd; it
6013 checks after acquiring the lock that its file descriptor still refers to
6014 the same named file.
6016 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6017 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6020 11. Added the smtp_active_hostname option. If used, this will typically be made
6021 to depend on the incoming interface address. Because $interface_address is
6022 not set up until the daemon has forked a reception process, error responses
6023 that can happen earlier (such as "too many connections") no longer contain
6026 12. If an expansion in a condition on a "warn" statement fails because a lookup
6027 defers, the "warn" statement is abandoned, and the next ACL statement is
6028 processed. Previously this caused the whole ACL to be aborted.
6030 13. Added the iplsearch lookup type.
6032 14. Added ident_timeout as a log selector.
6034 15. Added tls_certificate_verified as a log selector.
6036 16. Added a global option tls_require_ciphers (compare the smtp transport
6037 option of the same name). This controls incoming TLS connections.
6039 17. I finally figured out how to make tls_require_ciphers do a similar thing
6040 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6041 before starting the TLS session.
6043 18. Tabs are now shown as \t in -bP output.
6045 19. If the log selector return_path_on_delivery was set, Exim crashed when
6046 bouncing a message because it had too many Received: header lines.
6048 20. If two routers both had headers_remove settings, and the first one included
6049 a superfluous trailing colon, the final name in the first list and the
6050 first name in the second list were incorrectly joined into one item (with a
6051 colon in the middle).
6057 1. Added -C and -D options to the exinext utility, mainly to make it easier
6058 to include in the automated testing, but these could be helpful when
6059 multiple configurations are in use.
6061 2. The exinext utility was not formatting the output nicely when there was
6062 an alternate port involved in the retry record key, nor when there was a
6063 message id as well (for retries that were specific to a specific message
6064 and a specific host). It was also confused by IPv6 addresses, because of
6065 the additional colons they contain. I have fixed the IPv4 problem, and
6066 patched it up to do a reasonable job for IPv6.
6068 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6069 delivery, the log line now contains "pipelined" if PIPELINING was used.
6071 4. An SMTP transport process used to panic and die if the bind() call to set
6072 an explicit outgoing interface failed. This has been changed; it is now
6073 treated in the same way as a connect() failure.
6075 5. A reference to $sender_host_name in the part of a conditional expansion
6076 that was being skipped was still causing a DNS lookup. This no longer
6079 6. The def: expansion condition was not recognizing references to header lines
6080 that used bh_ and bheader_.
6082 7. Added the _cache feature to named lists.
6084 8. The code for checking quota_filecount in the appendfile transport was
6085 allowing one more file than it should have been.
6087 9. For compatibility with Sendmail, the command line option
6095 and sets the incoming protocol and host name (for trusted callers). The
6096 host name and its colon can be omitted when only the protocol is to be set.
6097 Note the Exim already has two private options, -pd and -ps, that refer to
6098 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6099 "s", but I don't think that's a major issue.
6101 10. A number of refactoring changes to the code, none of which should affect
6104 (a) The number of logging options was getting close to filling up the
6105 32-bit word that was used as a bit map. I have split them into two classes:
6106 those that are passed in the argument to log_write(), and those that are
6107 only ever tested independently outside of that function. These are now in
6108 separate 32-bit words, so there is plenty of room for expansion again.
6109 There is no change in the user interface or the logging behaviour.
6111 (b) When building, for example, log lines, the code previously used a
6112 macro that called string_cat() twice, in order to add two strings. This is
6113 not really sufficiently general. Furthermore, there was one instance where
6114 it was actually wrong because one of the argument was used twice, and in
6115 one call a function was used. (As it happened, calling the function twice
6116 did not affect the overall behaviour.) The macro has been replaced by a
6117 function that can join an arbitrary number of extra strings onto a growing
6120 (c) The code for expansion conditions now uses a table and a binary chop
6121 instead of a serial search (which was left over from when there were very
6122 few conditions). Also, it now recognizes conditions like "pam" even when
6123 the relevant support is not compiled in: a suitably worded error message is
6124 given if an attempt is made to use such a condition.
6126 11. Added ${time_interval:xxxxx}.
6128 12. A bug was causing one of the ddress fields not to be passed back correctly
6129 from remote delivery subprocesses. The field in question was not being
6130 subsequently used, so this caused to problems in practice.
6132 13. Added new log selectors queue_time and deliver_time.
6134 14. Might have fixed a bug in maildirsizefile handling that threw up
6135 "unexpected character" debug warnings, and recalculated the data
6136 unnecessarily. In any case, I expanded the warning message to give more
6139 15. Added the message "Restricted characters in address" to the statements in
6140 the default ACL that block characters like @ and % in local parts.
6142 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6143 Three changes have been made:
6145 (a) There was a serious bug; a negative response to MAIL caused the whole
6146 recipient domain to be cached as invalid, thereby blocking all messages
6147 to all local parts at the same domain, from all senders. This bug has
6148 been fixed. The domain is no longer cached after a negative response to
6149 MAIL if the sender used is not empty.
6151 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6154 (c) A new callout option, "use_sender" has been added for people who want
6155 the modified behaviour.
6161 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6164 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6165 indeed breaks things for older releases.
6167 3. Added additional logging to the case where there is a problem reading data
6168 from a filter that is running in a subprocess using a pipe, in order to
6169 try to track down a specific problem.
6171 4. Testing facility fudge: when running in the test harness and attempting
6172 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6173 getting "No route to host". Convert this to a timeout.
6175 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6178 6. Some OS don't have socklen_t but use size_t instead. This affects the
6179 fifth argument of getsockopt() amongst other things. This is now
6180 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6181 can be set for individual OS. I have set it for SunOS5, OSF1, and
6182 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6183 some earlier ones do not.
6185 7. Change 4.30/15 was not doing the test caselessly.
6187 8. The standard form for an IPv6 address literal was being rejected by address
6188 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6189 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6190 this, as well as the form without the "IPv6" on the front (but only when
6191 address literals are enabled, of course).
6193 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6195 10. Exim crashed if a message with an empty sender address specified by -f
6196 encountered a router with an errors_to setting. This could be provoked only
6197 by a command such as
6201 where an empty string was supplied; "<>" did not hit this bug.
6203 11. Installed PCRE release 4.5.
6205 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6206 remained set. It is now erased.
6208 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6209 times from message ids (which are base 36 rather than the normal 62).
6211 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6212 were being counted as actual protocol errors, and logged if the log
6213 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6214 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6215 and DATA following a set of rejected RCPTs do not count as protocol errors.
6216 In other words, Exim assumes they were pipelined, though this may not
6217 actually be the case. Of course, in all cases the client gets an
6218 appropriate error code.
6220 15. If a lookup fails in an ACL condition, a message about the failure may
6221 be available; it is used if testing the ACL cannot continue, because most
6222 such messages specify what the cause of the deferral is. However, some
6223 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6224 that caused an old message to be retained and used if a later statement
6225 caused a defer, replacing the real cause of the deferral.
6227 16. If an IP address had so many PTR records that the DNS lookup buffer
6228 was not large enough to hold them, Exim could crash while trying to process
6229 the truncated data. It now detects and logs this case.
6231 17. Further to 4.21/58, another change has been made: if (and only if) the
6232 first line of a message (the first header line) ends with CRLF, a bare LF
6233 in a subsequent header line has a space inserted after it, so as not to
6234 terminate the header.
6236 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6237 unnecessarily, used atoi() instead of strtol(), and didn't check the
6238 termination when getting file sizes from file names by regex.
6240 19. Completely re-implemented the support for maildirsize files, in the light
6241 of a number of problems with the previous contributed implementation
6242 (4.30/29). In particular:
6244 . If the quota is zero, the maildirsize file is maintained, but no quota is
6247 . If the maildir directory does not exist, it is created before any attempt
6248 to write a maildirsize file.
6250 . The quota value in the file is just a cache; if the quota is changed in
6251 the transport, the new value overrides.
6253 . A regular expression is available for excluding directories from the
6256 20. The autoreply transport checks the characters in options that define the
6257 message's headers; it allows continued headers, but it was checking with
6258 isspace() after an embedded newline instead of explicitly looking for a
6261 21. If all the "regular" hosts to which an address was routed had passed their
6262 expiry times, and had not reached their retry times, the address was
6263 bounced, even if fallback hosts were defined. Now Exim should go on to try
6266 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6267 equivalent code in the SMTP transport. Some hosts send humungous responses
6268 to HELO/EHLO, more than 1024 it seems.
6270 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6271 gives compiler warnings in some environments. I've now done it "properly",
6274 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6275 (because of problems with the built-in one) was declared to return uschar *
6276 instead of char *, causing compiler failure.
6278 25. Fixed a file descriptor leak when processing alias/forward files.
6280 26. Fixed a minor format string issue in dbfn.c.
6282 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6284 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6285 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6288 29. When a sender address is verified, it is cached, to save repeating the test
6289 when there is more than one recipient in a message. However, when the
6290 verification involves a callout, it is possible for different callout
6291 options to be set for different recipients. It is too complicated to keep
6292 track of this in the cache, so now Exim always runs a verification when a
6293 callout is required, relying on the callout cache for the optimization.
6294 The overhead is duplication of the address routing, but this should not be
6297 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6298 to be verified with callout=postmaster, and the main callout worked but the
6299 postmaster check failed, the verification correctly failed. However, if a
6300 subsequent RCPT command asked for sender verification *without* the
6301 postmaster check, incorrect caching caused this verification also to fail,
6304 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6305 it was not caching the DNS options (qualify_single, search_parents) that
6306 were used when the lookup failed. A subsequent lookup with different
6307 options therefore always gave the same answer, though there were cases
6308 where it should not have. (Example: a "domains = !$mx_any" option on a
6309 dnslookup router: the "domains" option is always processed without any
6310 widening, but the router might have qualify_single set.) Now Exim uses the
6311 cached value only when the same options are set.
6313 32. Added John Jetmore's "exipick" utility to the distribution.
6315 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6316 than a timeout (e.g. a certificate is required, and is not provided), an
6317 Exim server now closes the connection immediately. Previously it waited for
6318 the client to close - but if the client is SSL, it seems that they each
6319 wait for each other, leading to a delay before one of them times out.
6321 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6322 maintained 0.8.x compatibility because I don't think many are using it, and
6323 it is clearly obsolete.
6325 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6328 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6329 client certificate was expired. A simple patch fixes this, though I don't
6330 understand the full logic of why the verify callback is called multiple
6333 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6334 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6335 which causes problems with some clients (such as the Certicom SSL Plus
6336 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6337 disables the coutermeasure allowing Eudora to connect."
6339 38. Exim was not checking that a write() to a log file succeeded. This could
6340 lead to Bad Things if a log got too big, in particular if it hit a file
6341 size limit. Exim now panics and dies if it cannot write to a log file, just
6342 as it does if it cannot open a log file.
6344 39. Modified OS/Makefile-Linux so that it now contains
6346 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6348 The two -D definitions ensure that Exim is compiled with large file
6349 support, which makes it possible to handle log files that are bigger than
6352 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6353 instance) a domain was checked against a named list that involved a lookup,
6354 causing $domain_data to be set, then another domain was checked against the
6355 same list, then the first domain was re-checked, the value of $domain_data
6356 after the final check could be wrong. In particular, if the second check
6357 failed, it could be set empty. This bug probably also applied to
6360 41. The strip_trailing_dot option was not being applied to the address given
6361 with the -f command-line option.
6363 42. The code for reading a message's header from the spool was incrementing
6364 $received_count, but never initializing it. This meant that the value was
6365 incorrect (doubled) while delivering a message in the same process in which
6366 it was received. In the most common configuration of Exim, this never
6367 happens - a fresh exec is done - but it can happen when
6368 deliver_drop_privilege is set.
6370 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6371 it now includes up to 150 characters of the unexpected data in the log
6374 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6375 and building data strings. The size of both of these buffers was 10 000
6376 bytes - far larger than anybody would *ever* want, thought I. Needless to
6377 say, somebody hit the limit. I have increased the maximum line length to
6378 20 000 and the maximum data length of concatenated lines to 100 000. I have
6379 also fixed two bugs, because there was no checking on these buffers. Tsk,
6380 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6381 buffer is too small.
6383 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6384 lsearch lookups. Now it does.
6386 46. When parsing a route_list item in a manualroute router, a fixed-length
6387 buffer was used for the list of hosts. I made this 1024 bytes long,
6388 thinking that nobody would ever have a list of hosts that long. Wrong.
6389 Somebody had a whole pile of complicated expansion conditions, and the
6390 string was silently truncated, leading to an expansion error. It turns out
6391 that it is easier to change to an unlimited length (owing to other changes
6392 that have happened since this code was originally written) than to build
6393 structure for giving a limitation error. The length of the item that
6394 expands into the list of hosts is now unlimited.
6396 47. The lsearch lookup could not handle data where the length of text line was
6397 more than 4095 characters. Such lines were truncated, leading to shortened
6398 data being returned. It should now handle lines of any length.
6400 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6401 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6404 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6405 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6406 no longer happens in comment lines. A list of replacements is now placed
6407 at the head of all of the source files, except those whose only change is
6408 to replace PERL_COMMAND in the very first #! line.
6410 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6411 messages on the queue, with a bottom-up merge sort, using code contributed
6412 by Michael Haardt. This should make operations like -bp somewhat faster on
6413 large queues. It won't affect queue runners, except when queue_run_in_order
6416 51. Installed eximstats 1.31 in the distribution.
6418 52. Added support for SRV lookups to the dnslookup router.
6420 53. If an ACL referred to $message_body or $message_body_end, the value was not
6421 reset for any messages that followed in the same SMTP session.
6423 54. The store-handling optimization for building very long strings was not
6424 differentiating between the different store pools. I don't think this
6425 actually made any difference in practice, but I've tidied it.
6427 55. While running the routers to verify a sender address, $sender_address
6428 was still set to the sender address. This is wrong, because when routing to
6429 send a bounce to the sender, it would be empty. Therefore, I have changed
6430 it so that, while verifying a sender address, $sender_address is set to <>.
6431 (There is no change to what happens when verifying a recipient address.)
6433 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6434 target A or AAAA records (if not already returned) without resetting the
6435 qualify_single or search_parents options of the DNS resolver. These are
6436 inappropriate in this case because the targets of MX and SRV records must
6437 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6438 target that, when qualified, matched something in the local domain. These
6439 two options are now turned off when doing these lookups.
6441 57. It seems that at least some releases of Reiserfs (which does not have the
6442 concept of a fixed number of inodes) returns zero and not -1 for the
6443 number of available inodes. This interacted badly with check_spool_inodes,
6444 which assumed that -1 was the "no such thing" setting. What I have done is
6445 to check that the total number of inodes is greater than zero before doing
6446 the test of how many are available.
6448 58. When a "warn" ACL statement has a log_message modifier, the message is
6449 remembered, and not repeated. This is to avoid a lot of repetition when a
6450 message has many recipients that cause the same warning to be written.
6451 However, Exim was preserving the list of already written lines for an
6452 entire SMTP session, which doesn't seem right. The memory is now reset if a
6453 new message is started.
6455 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6456 debugging output unless log_rewrite was also set.
6458 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6459 of (int)(handle) when we know that handle contains (void *)(-1).
6461 61. The Exim daemon panic-logs an error return when it closes the incoming
6462 connection. However "connection reset by peer" seems to be common, and
6463 isn't really an error worthy of noting specially, so that particular error
6466 62. When Exim is trying to find all the local interfaces, it used to panic and
6467 die if the ioctl to get the interface flags failed. However, it seems that
6468 on at least one OS (Solaris 9) it is possible to have an interface that is
6469 included in the list of interfaces, but for which you get a failure error
6470 for this call. This happens when the interface is not "plumbed" into a
6471 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6472 failure of the "get flags" call assumes that the interface is down.
6474 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6475 makes life easier for people who are doing arithmetic on fields extracted
6476 from dates, where you often get leading zeros that should not be
6477 interpreted as octal.
6479 64. Added qualify_domain to the redirect router, to override the global
6482 65. If a pathologically long header line contained very many addresses (the
6483 report of this problem mentioned 10 000) and each of them was rewritten,
6484 Exim could use up a very large amount of memory. (It kept on making new
6485 copies of the header line as it rewrote, and never released the old ones.)
6486 At the expense of a bit more processing, the header rewriting function has
6487 been changed so that it no longer eats memory in this way.
6489 66. The generation of the Received: header has been moved from the time that a
6490 message starts to be received, to the time that it finishes. The timestamp
6491 in the Received: header should now be very close to that of the <= log
6492 line. There are two side-effects of this change:
6494 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6495 logged header lines no longer include the local Received: line, because
6496 it has not yet been created. The same applies to a copy of the message
6497 that is returned to a non-SMTP sender when a message is rejected.
6499 (b) When a filter file is tested using -bf, no additional Received: header
6500 is added to the test message. After some thought, I decided that this
6503 This change does not affect the value of $received_for. It is still set
6504 after address rewriting, but before local_scan() is called.
6506 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6508 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6509 gave an unhelpful panic error message, and a defer error. I have managed to
6510 change this behaviour so that it now rejects any supplied certificate,
6511 which seems right, as the list of acceptable certificates is empty.
6513 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6514 gave an unhelpful defer error. I have not managed to make this reject any
6515 supplied certificates, but the error message it gives is "no certificate
6516 supplied", which is not helpful.
6518 70. exigrep's output now also includes lines that are not associated with any
6519 message, but which match the given pattern. Implemented by a patch from
6520 Martin Sluka, which also tidied up the Perl a bit.
6522 71. Recipient callout verification, like sender verification, was using <> in
6523 the MAIL FROM command. This isn't really the right thing, since the actual
6524 sender may affect whether the remote host accepts the recipient or not. I
6525 have changed it to use the actual sender in the callout; this means that
6526 the cache record is now keyed on a recipient/sender pair, not just the
6527 recipient address. There doesn't seem to be a real danger of callout loops,
6528 since a callout by the remote host to check the sender would use <>.
6529 [SEE ABOVE: changed after hitting problems.]
6531 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6532 temporary errors. However, in the case of such a code being given after
6533 the end of a data transmission (i.e. after ".") Exim was failing to write
6534 a retry record for the message. (Yes, there was some broken host that was
6535 actually sending 8xx at this point.)
6537 73. An unknown lookup type in a host list could cause Exim to panic-die when
6538 the list was checked. (An example that provoked this was putting <; in the
6539 middle of a list instead of at the start.) If this happened during a DATA
6540 ACL check, a -D file could be left lying around. This kind of configuration
6541 error no longer causes Exim to die; instead it causes a defer error. The
6542 incident is still logged to the main and panic logs.
6544 74. Buglet left over from Exim 3 conversion. The message "too many messages
6545 in one connection" was written to the rejectlog but not the mainlog, except
6546 when address rewriting (yes!) was being logged.
6548 75. Added write_rejectlog option.
6550 76. When a system filter was run not as root (that is, when system_filter_user
6551 was set), the values of the $n variables were not being returned to the
6552 main process; thus, they were not subsequently available in the $sn
6555 77. Added +return_path_on_delivery log selector.
6557 78. A connection timeout was being treated differently from recipients deferred
6558 when testing hosts_max_try with a message that was older than the host's
6559 retry timeout. (The host should not be counted, thus allowing all hosts to
6560 be tried at least once before bouncing.) This may have been the cause of an
6561 occasionally reported bug whereby a message would remain on the queue
6562 longer than the retry timeout, but would be bounced if a delivery was
6563 forced. I say "may" because I never totally pinned down the problem;
6564 setting up timeout/retry tests is difficult. See also the next item.
6566 79. The ultimate address timeout was not being applied to errors that involved
6567 a combination of host plus message (for example, a timeout on a MAIL
6568 command). When an address resolved to a number of possible hosts, and they
6569 were not all tried for each delivery (e.g. because of hosts_max_try), a
6570 message could remain on the queue longer than the retry timeout.
6572 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6575 81. Fixed an obscure SMTP outgoing bug which required at least the following
6576 conditions: (a) there was another message waiting for the same server;
6577 (b) the server returned 5xx to all RCPT commands in the first message so
6578 that the message was not completed; (c) the server dropped the connection
6579 or gave a negative response to the RSET that Exim sends to abort the
6580 transaction. The observed case was a dropped connection after DATA that had
6581 been sent in pipelining mode. That is, the server had advertised PIPELINING
6582 but was not implementing it correctly. The effect of the bug was incorrect
6583 behaviour, such as trying another host, and this could lead to a crash.
6589 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6590 and daemon.c were passed as pointers to ints; they should have been
6591 pointers to socklen_t variables (which are typically unsigned ints).
6593 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6596 3. Fixed a really odd bug that affected only the testing scheme; patching a
6597 certain fixed string in the binary changed the value of another string that
6598 happened to be identical to the end of the original first string.
6600 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6601 name", it returns that address as the IP address. On some operating
6602 systems (e.g. Solaris), it also passes back the IP address string as the
6603 "host name". However, on others (e.g. Linux), it passes back an empty
6604 string. Exim wasn't checking for this, and was changing the host name to an
6605 empty string, assuming it had been canonicalized.
6607 5. Although rare, it is permitted to have more than one PTR record for a given
6608 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6609 all the names associated with an address, because they do in Solaris.
6610 However, it seems that they do not in Linux for data that comes from the
6611 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6612 I found this out when I moved to a new Linux workstation and tried to run
6613 the Exim test suite.
6615 To get round this problem I have changed the code so that it now does its
6616 own call to the DNS to look up PTR records when searching for a host name.
6617 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6618 addresses that are only in /etc/hosts are still found.
6620 This behaviour is, however, controlled by an option called host_lookup_
6621 order, which defaults to "bydns:byaddr". If people want to use the other
6622 order, or indeed, just use one or the other means of lookup, they can
6623 specify it in this variable.
6625 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6626 some operating systems, this comes back from gethostbyaddr() as an empty
6627 string, and this is what Exim used to test for. However, it seems that in
6628 other systems, "." is yielded. Exim now tests for this case too.
6630 7. The values of check_spool_space and check_log_space are now held internally
6631 as a number of kilobytes instead of an absolute number of bytes. If a
6632 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6633 kilobyte. This means that much larger values can be stored.
6635 8. Exim monitor: an attempt to get the action menu when not actually pointing
6636 at a message produces an empty menu entitled "No message selected". This
6637 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6638 no entries in it ("Shell widget menu has zero width and/or height"). So I
6639 have added a single, blank menu entry in this case.
6641 9. Added ${quote_local_part.
6643 10. MIME decoding is now applied to the contents of Subject: header lines when
6646 11. Now that a reference to $sender_host_address automatically causes a reverse
6647 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6648 host lookup before query-style lookups in lists that might use this
6649 variable. This has therefore been abolished, and the "net-" prefix is no
6650 longer necessary for query-style lookups.
6652 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6653 have been a typo for LFLAGS, so it has been changed.
6655 13. The install script calls Exim with "-C /dev/null" in order to find the
6656 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6657 to be output. However, since Exim outputs its version number before the
6658 error, it didn't break the script. It just looked ugly. I fixed this by
6659 always allowing "-C /dev/null" if the caller is root.
6661 14. Ignore overlarge ACL variable number when reading spool file - insurance
6662 against a later release with more variables having written the file.
6664 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6665 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6666 this, as well as the form without the "IPv6" on the front.
6668 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6669 OS/Makefile-Darwin file.
6671 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6672 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6674 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6675 variable. However, some OS use ints for inodes. I've added cast to long int
6676 to get rid of the compiler warning.
6678 19. I had forgotten to lock out "/../" in configuration file names when
6679 ALT_CONFIG_PREFIX was set.
6681 20. Routers used for verification do not need to specify transports. However,
6682 if such a router generated a host list, and callout was configured, Exim
6683 crashed, because it could not find a port number from the (non-existent)
6684 transport. It now assumes port 25 in this circumstance.
6686 21. Added the -t option to exigrep.
6688 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6689 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6690 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6691 them is set, LOOKUP_LSEARCH is forced.
6693 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6695 24. Added sender and host information to the "rejected by local_scan()" log
6696 line; previously there was no indication of these.
6698 25. Added .include_if_exists.
6700 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6701 Linux. It turns out that not all file systems support this. Apparently some
6702 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6703 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6704 not supported on the file descriptor, is now ignored when Exim is trying to
6705 sync a directory. This applies only to Linux.
6707 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6709 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6710 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6713 29. Added support for maildirsize files from supplied patch (modified a bit).
6715 30. The use of :fail: followed by an empty string could lead Exim to respond to
6716 sender verification failures with (e.g.):
6718 550 Verification failed for <xxx>
6719 550 Sender verify failed
6721 where the first response line was missing the '-' that indicates it is not
6722 the final line of the response.
6724 31. The loop for finding the name of the user that called Exim had a hardwired
6725 limit of 10; it now uses the value of finduser_retries, which is used for
6726 all other user lookups.
6728 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6731 33. Exim was neglecting to zero errno before one call of strtol() when
6732 expanding a string and expecting an integer value. On some systems this
6733 resulted in spurious "integer overflow" errors. Also, it was casting the
6734 result into an int without checking.
6736 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6737 did not work. The code looks as if it has *never* worked, though it appears
6738 to have been documented since at least release 1.62. I have made it work.
6740 35. The "timeout_DNS" error in retry rules, also documented since at least
6741 1.62, also never worked. As it isn't clear exactly what this means, and
6742 clearly it isn't a major issue, I have abolished the feature by treating it
6743 as "timeout", and writing a warning to the main and panic logs.
6745 36. The display of retry rules for -brt wasn't always showing the error code
6748 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6749 timeout_connect_A, timeout_connect_MX.
6751 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6752 to the empty sender.
6754 39. The daemon was not analysing the content of -oX till after it had closed
6755 stderr and disconnected from the controlling terminal. This meant that any
6756 syntax errors were only noted on the panic log, and the return code from
6757 the command was 0. By re-arranging the code a little, I've made the
6758 decoding happen first, so such errors now appear on stderr, and the return
6759 code is 1. However, the actual setting up of the sockets still happens in
6760 the disconnected process, so errors there are still only recorded on the
6763 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6764 connections (as happens on some IP stacks) was logged at start up time as
6765 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6766 it from "IPv6 and IPv4", which means that two separate sockets are being
6769 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6770 says whether AF_INET or AF_INET6 was passed as an argument.
6772 42. Exiwhat output was messed up when time zones were included in log
6775 43. Exiwhat now gives more information about the daemon's listening ports,
6776 and whether -tls-on-connect was used.
6778 44. The "port" option of the smtp transport is now expanded.
6780 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6781 silently ignored. Now an error message is written to the main and panic
6784 46. There's a new ACL modifier called "logwrite" which writes to a log file
6785 as soon as it is encountered.
6787 47. Added $local_user_uid and $local_user_gid at routing time.
6789 48. Exim crashed when trying to verify a sender address that was being
6792 49. Exim was recognizing only a space character after ".include". It now also
6793 recognizes a tab character.
6795 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6796 extracting the relevant information from the specification. The man page no
6797 longer contains scrambled data for the -d option, and I've added a section
6798 at the front about calling Exim under different names.
6800 51. Added "extra_headers" argument to the "mail" command in filter files.
6802 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6805 53. Installed eximstats 1.29.
6807 54. Added transport_filter_timeout as a generic transport option.
6809 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6810 Cc: header lines. This was required by RFC 822, but it not required by RFC
6813 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6814 incoming messages that did not have them. Now it does so only if the
6815 message originates locally, that is, if there is no associated remote host
6816 address. When Resent- header lines are present, this applies to the Resent-
6817 lines rather than the non-Resent- lines.
6819 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6820 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6822 58. Messages for configuration errors now include the name of the main
6823 configuration file - useful now that there may be more than one file in a
6824 list (.included file names were always shown).
6826 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6827 for those rare installations that do not start the daemon as root or run it
6828 setuid root. I've cut out the call to initgroups() if the daemon is not
6831 60. The Exim user and group can now be bound into the binary as text strings
6832 that are looked up at the start of Exim's processing.
6834 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6836 62. Added $mailstore_basename variable.
6838 63. Installed patch to sieve.c from Michael Haardt.
6840 64. When Exim failed to open the panic log after failing to open the main log,
6841 the original message it was trying to log was written to stderr and debug
6842 output, but if they were not available (the usual case in production), it
6843 was lost. Now it is written to syslog before the two lines that record the
6844 failures to open the logs.
6846 65. Users' Exim filters run in subprocesses under the user's uid. It is
6847 possible for a "deliver" command or an alias in a "personal" command to
6848 provoke an address rewrite. If logging of address rewriting is configured,
6849 this fails because the process is not running as root or exim. There may be
6850 a better way of dealing with this, but for the moment (because 4.30 needs
6851 to be released), I have disabled address rewrite logging when running a
6852 filter in a non-root, non-exim process.
6858 1. The buildconfig auxiliary program wasn't quoting the value set for
6859 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6860 not defined. This bug was masked in 4.22 by the effect that was fixed in
6863 2. Some messages that were rejected after a message id was allocated were
6864 shown as "incomplete" by exigrep. It no longer does this for messages that
6865 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6867 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6868 have allow_domain_literals set, the ID did not get logged in the <= line.
6869 Domain literals are now always recognized in Message-ID: header lines.
6871 4. The first argument for a ${extract expansion item is the key name or field
6872 number. Leading and trailing spaces in this item were not being ignored,
6873 causing some misleading effects.
6875 5. When deliver_drop_privilege was set, single queue runner processes started
6876 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6877 same command in the process it spins off) were not dropping privilege.
6879 6. When the daemon running as "exim" started a queue runner, it always
6880 re-executed Exim in the spun-off process. This is a waste of effort when
6881 deliver_drop_privilege is set. The new process now just calls the
6882 queue-runner function directly.
6888 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6891 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6892 ignored. Though the use of -oP was forcing the writing of a pid file, it
6893 was always written to the default place.
6895 3. If the message "no IP address found for host xxxx" is generated during
6896 incoming verification, it is now followed by identification of the incoming
6897 connection (so you can more easily find what provoked it).
6899 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6901 5. Added some features to "harden" Exim a bit more against certain attacks:
6903 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6904 be put in Local/Makefile. This is like the never_users runtime option,
6905 but it cannot be overridden. The default setting is "root".
6907 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6908 prefix string with which any file named in a -C command line option
6911 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6912 is retained for -C and -D only if the caller of Exim is root. Without
6913 it, the exim user may also use -C and -D and retain privilege.
6915 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6916 command line option is disabled.
6918 6. Macro names set by the -D option must start with an upper case letter, just
6919 like macro names defined in the configuration file.
6921 7. Added "dereference=" facility to LDAP.
6923 8. Two instances of the typo "uknown" in the source files are fixed.
6925 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6926 the Configure-Makefile script screwed up while processing it.
6928 10. Incorporated PCRE 4.4.
6930 11. The SMTP synchronization check was not operating right at the start of an
6931 SMTP session. For example, it could not catch a HELO sent before the client
6932 waited for the greeting. There is now a check for outstanding input at the
6933 point when the greeting is written. Because of the duplex, asynchronous
6934 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6935 way, but not yet received, when the check is performed.
6937 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6938 on TCP/IP sockets, because this apparently causes some broken clients to
6941 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6942 unchanged) from the Cygwin maintainer.
6944 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6945 was supported instead of testing for mbx. Effectively a typo.
6947 15. The spa authenticator server code was not checking that the input it
6948 received was valid base64.
6950 16. The debug output line for the "set" modifier in ACLs was not showing the
6951 name of the variable that was being set.
6953 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6955 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6956 lookup on its own. Something else was needed to trigger the lookup. For
6957 example, a match in host_lookup or the need for a host name in a host list.
6958 Now, if $sender_host_name is referenced and the host name has not yet been
6959 looked up, a lookup is performed. If the lookup fails, the variable remains
6960 empty, and $host_lookup_failed is set to "1".
6962 19. Added "eqi" as a case-independent comparison operator.
6964 20. The saslauthd authentication condition could segfault if neither service
6965 nor realm was specified.
6967 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6968 error message that was logged was misleading, and incoming SMTP
6969 connections were dropped. The message is now more accurate, and temporary
6970 errors are given to SMTP connections.
6972 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6973 (except RSET) with a 421 error, until QUIT is received. However, it was
6974 failing to send a response to QUIT.
6976 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6977 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6978 ACL is now run after the helo_try_verify_hosts code.
6980 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6981 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6982 case-independent, so other case variants are also recognized). Apparently
6983 some systems use these upper case variants.
6985 25. If more than two messages were waiting for the same host, and a transport
6986 filter was specified for the transport, Exim sent two messages over the
6987 same TCP/IP connection, and then failed with "socket operation on non-
6988 socket" when it tried to send the third.
6990 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6992 27. The extern definition of crypt16() in expand.c was not being excluded when
6993 the OS had its own crypt16() function.
6995 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6996 as a preferred synonym for return_size_limit, both as an option and as an
6999 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7001 30. Changed the default configuration ACL to relax the local part checking rule
7002 for addresses that are not in any local domains. For these addresses,
7003 slashes and pipe symbols are allowed within local parts, but the sequence
7004 /../ is explicitly forbidden.
7006 31. SPA server authentication was not clearing the challenge buffer before
7009 32. log_message in a "warn" ACL statement was writing to the reject log as
7010 well as to the main log, which contradicts the documentation and doesn't
7011 seem right (because no rejection is happening). So I have stopped it.
7013 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7014 However, I am unable to do any testing of this.
7016 34. Fixed an infelicity in the appendfile transport. When checking directories
7017 for a mailbox, to see if any needed to be created, it was accidentally
7018 using path names with one or more superfluous leading slashes; tracing
7019 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7021 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7022 contents are added to the log line that is written for every discarded
7023 recipient. (Previously a log_message setting was ignored.)
7025 36. The ${quote: operator now quotes the string if it is empty.
7027 37. The install script runs exim in order to find its version number. If for
7028 some reason other than non-existence or emptiness, which it checks, it
7029 could not run './exim', it was installing it with an empty version number,
7030 i.e. as "exim-". This error state is now caught, and the installation is
7033 38. An argument was missing from the function that creates an error message
7034 when Exim fails to connect to the socket for saslauthd authentication.
7035 This could cause Exim to crash, or give a corrupted message.
7037 39. Added isip, isip4, and isip6 to ${if conditions.
7039 40. The ACL variables $acl_xx are now saved with the message, and can be
7040 accessed later in routers, transports, and filters.
7042 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7043 strings in the file are not string-expanded.
7045 42. If a MAIL command specified a SIZE value that was too large to fit into an
7046 int variable, the check against message_size_limit failed. Such values are
7047 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7048 day this will have to be increased, but I don't think I want to be around
7049 when emails are that large.
7056 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7057 iconv() is not standard in FreeBSD.
7059 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7060 IPv6 enabled. The observed symptom was a segmentation fault on return from
7061 the function os_common_find_running_interfaces() in src/os.c.
7063 3. In the check_special_case() function in daemon.c I had used "errno" as an
7064 argument name, which causes warnings on some systems. This was basically a
7065 typo, since it was named "eno" in the comments!
7067 4. The code that waits for the clock to tick (at a resolution of some fraction
7068 of a second) so as to ensure message-id uniqueness was always waiting for
7069 at least one whole tick, when it could have waited for less. [This is
7070 almost certainly not relevant at current processor speeds, where it is
7071 unlikely to ever wait at all. But we try to future-proof.]
7073 5. The function that sleeps for a time interval that includes fractions of a
7074 second contained a race. It did not block SIGALRM between setting the
7075 timer, and suspending (a couple of lines later). If the interval was short
7076 and the sigsuspend() was delayed until after it had expired, the suspension
7077 never ended. On busy systems this could lead to processes getting stuck for
7080 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7081 process, before it forks any delivery processes. The open lookup caching
7082 mechanism meant that the open file or database connection was passed into
7083 the delivery process. The problem was that delivery processes always tidy
7084 up cached lookup data. This could cause a problem for the next delivery
7085 process started by the queue runner, because the external queue runner
7086 process does not know about the closure. So the next delivery process
7087 still has data in the lookup cache. In the case of a file lookup, there was
7088 no problem because closing a file descriptor in a subprocess doesn't affect
7089 the parent. However, if the lookup was caching a connection to a database,
7090 the connection was closed, and the second delivery process was likely to
7091 see errors such as "PGSQL: query failed: server closed the connection
7092 unexpectedly". The problem has been fixed by closing all cached lookups
7093 in a queue runner before running a delivery process.
7095 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7096 seem to have the "const" qualifier which it has on other OS. I've
7099 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7100 *of the same type* (client or server) with the same public name that an
7101 error should be diagnosed.
7103 9. When Exim looked up a host name for an IP address, but failed to find the
7104 original IP address when looking up the host name (a safety check), it
7105 output the message "<ip address> does not match any IP for NULL", which was
7106 confusing, to say the least. The bug was that the host name should have
7107 appeared instead of "NULL".
7109 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7110 user that is built into the binary, and the -C or -D options is used, root
7111 privilege is dropped before the configuration file is read. In addition,
7112 logging is switched to stderr instead of the normal log files. If the
7113 configuration then re-defines the Exim user, the unprivileged environment
7114 is probably not what is expected, so Exim logs a panic warning message (but
7117 However, if deliver_drop_privilege is set, the unprivileged state may well
7118 be exactly what is intended, so the warning has been cut out in that case,
7119 and Exim is allowed to try to write to its normal log files.
7125 1. smtp_return_error_details was not giving details for temporary sender
7126 or receiver verification errors.
7128 2. Diagnose a configuration error if two authenticators have the same public
7131 3. Exim used not to create the message log file for a message until the first
7132 delivery attempt. This could be confusing when incoming messages were held
7133 for policy or load reasons. The message log file is now created at the time
7134 the message is received, and an initial "Received" line is written to it.
7136 4. The automatically generated man page for command line options had a minor
7137 bug that caused no ill effects; however, a more serious problem was that
7138 the procedure for building the man page automatically didn't always
7139 operate. Consequently, release 4.20 contains an out-of-date version. This
7140 shouldn't happen again.
7142 5. When building Exim with embedded Perl support, the script that builds the
7143 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7144 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7146 6. The freeze_tell option was not being used for messages that were frozen on
7147 arrival, either by an ACL or by local_scan().
7149 7. Added the smtp_incomplete_transaction log selector.
7151 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7152 was accepting AUTH without a new EHLO.
7154 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7155 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7158 10. Logging of TCP/IP connections (when configured) now happens in the main
7159 daemon process instead of the child process, so that the TCP/IP connection
7160 count is more accurate (but it can never be perfect).
7162 11. The use of "drop" in a nested ACL was not being handled correctly in the
7163 outer ACL. Now, if condition failure induced by the nested "drop" causes
7164 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7165 or "require"), the connection is dropped.
7167 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7168 that yield "discard" can now be used with an "accept" or a "discard" verb,
7169 but an error is generated for any others (because I can't see a useful way
7170 to define what should happen).
7172 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7173 are now processed in the same way as lines in the Exim configuration file.
7174 In particular, continuation lines are supported.
7176 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7178 15. Added -ti meaning -t -i.
7180 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7181 domains, and warn by logging if others are found.
7183 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7184 returned from GIFCONF when Exim is trying to find the list of interfaces on
7185 a host. The code in os.c has been modified to copy each ifreq to an aligned
7186 structure in all cases.
7188 Also, in some cases, the returned ifreq's were being copied to a 'struct
7189 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7190 means the last couple of bytes of an IPv6 address could be chopped if the
7191 ifreq contained only a normal sockaddr (14 bytes storage).
7193 18. Named domain lists were not supported in the hosts_treat_as_local option.
7194 An entry such as +xxxx was not recognized, and was treated as a literal
7197 19. Ensure that header lines added by a DATA ACL are included in the reject log
7198 if the ACL subsequently rejects the message.
7200 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7201 MD5 (which is deprecated).
7203 21. When testing a filter file using -bf, Exim was writing a message when it
7204 took the sender from a "From " line in the message, but it was not doing so
7205 when it took $return_path from a Return-Path: header line. It now does.
7207 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7208 with a valid header line field name (a series of printing characters
7209 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7211 23. Changed "disc" in the source to "disk" to conform to the documentation and
7212 the book and for uniformity.
7214 24. Ignore Sendmail's -Ooption=value command line item.
7216 25. When execve() failed while trying to run a command in a pipe transport,
7217 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7218 could be confused with a return value of 69 from the command itself. This
7219 has been changed to 127, the value the shell returns if it is asked to run
7220 a non-existent command. The wording for the related log line suggests a
7221 non-existent command as the problem.
7223 26. If received_header_text expands to an empty string, do not add a Received:
7224 header line to the message. (Well, it adds a token one on the spool, but
7225 marks it "old" so that it doesn't get used or transmitted.)
7227 27. Installed eximstats 1.28 (addition of -nt option).
7229 28. There was no check for failure on the call to getsockname() in the daemon
7230 code. This can fail if there is a shortage of resources on the system, with
7231 ENOMEM, for example. A temporary error is now given on failure.
7233 29. Contrary to the C standard, it seems that in some environments, the
7234 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7235 program. Exim now does this explicitly; it affects the formatting of
7236 timestamps using strftime().
7238 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7239 complaints. I've now initialized all the variables, to avoid this.
7241 32. Header lines added by a system filter were not being "seen" during
7242 transport-time rewrites.
7244 33. The info_callback() function passed to OpenSSL is set up with type void
7245 (*)(SSL *, int, int), as described somewhere. However, when calling the
7246 function (actually a macro) that sets it up, the type void(*)() is
7247 expected. I've put in a cast to prevent warnings from picky compilers.
7249 34. If a DNS black list lookup found a CNAME record, but there were no A
7250 records associated with the domain it pointed at, Exim crashed.
7252 35. If a DNS black list lookup returned more than one A record, Exim ignored
7253 all but the first. It now scans all returned addresses if a particular IP
7254 value is being sought. In this situation, the contents of the
7255 $dnslist_value variable are a list of all the addresses, separated by a
7258 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7259 to accept a host name and all its aliases if the forward lookup for any of
7260 them yielded the IP address of the incoming connection. Now it accepts only
7261 those names whose forward lookup yields the correct IP address. Any other
7262 names are discarded. This closes a loophole whereby a rogue DNS
7263 administrator could create reverse DNS records to break through a
7264 wildcarded host restriction in an ACL.
7266 37. If a user filter or a system filter that ran in a subprocess used any of
7267 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7268 the wrong values were passed to the pipe command ($thisaddress had the
7269 value of $0, $0 had the value of $1, etc). This bug was introduced by
7270 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7272 38. Improved the line breaking for long SMTP error messages from ACLs.
7273 Previously, if there was no break point between 40 and 75 characters, Exim
7274 left the rest of the message alone. Two changes have been made: (a) I've
7275 reduced the minimum length to 35 characters; (b) if it can't find a break
7276 point between 35 and 75 characters, it looks ahead and uses the first one
7277 that it finds. This may give the occasional overlong line, but at least the
7278 remaining text gets split now.
7280 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7281 file descriptors might be low, and that setting 1000 would always raise it.
7282 It turns out that in some environments, the limit is already over 1000 and
7283 that lowering it causes trouble. So now Exim takes care not to decrease it.
7285 40. When delivering a message, the value of $return_path is set to $sender_
7286 address at the start of routing (routers may change the value). By an
7287 oversight, this default was not being set up when an address was tested by
7288 -bt or -bv, which affected the outcome if any router or filter referred to
7291 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7292 log only when "message" or "log_message" are set. However, if one of the
7293 conditions was an address verification, or a call to a nested ACL, the
7294 messages generated by the underlying test were being passed through. This
7295 no longer happens. The underlying message is available in $acl_verify_
7296 message for both "message" and "log_message" expansions, so it can be
7297 passed through if needed.
7299 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7300 new expansion $bh_ to give the encoded byte string without charset
7301 translation. Translation happens only if iconv() is available; HAVE_ICONV
7302 indicates this at build time. HEADERS_CHARSET gives the charset to
7303 translate to; headers_charset can change it in the configuration, and
7304 "headers charset" can change it in an individual filter file.
7306 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7307 that creates RFC 2047 encoded "words" labels them as that charset instead
7308 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7309 expansion operator; (ii) when Exim creates a From: line for a local
7310 message; (iii) when a header line is rewritten to include a "phrase" part.
7312 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7313 buggy, causing it to skip the first lines of messages whose message ID
7314 ended in 'D'. This would not have bitten before Exim release 4.14, because
7315 message IDs were unlikely to end in 'D' before then. The effect was to have
7316 incorrect size information for certain domains.
7318 45. #include "config.h" was missing at the start of the crypt16.c module. This
7319 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7322 46. If there was a timeout during a "random" callout check, Exim treated it as
7323 a failure of the random address, and carried on sending RSET and the real
7324 address. If the delay was just some slowness somewhere, the response to the
7325 original RCPT would be taken as a response to RSET and so on, causing
7326 mayhem of various kinds.
7328 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7329 when I implemented it. It didn't allow for the fact that some option values
7330 may legitimately be negative (e.g. size_addition), and it didn't even do
7331 the right test for positive values.
7333 48. Domain names in DNS records are case-independent. Exim always looks them up
7334 in lower case. Some resolvers return domain names in exactly the case they
7335 appear in the zone file, that is, they may contain uppercase letters. Not
7336 all resolvers do this - some return always lower case. Exim was treating a
7337 change of case by a resolver as a change of domain, similar to a widening
7338 of a domain abbreviation. This triggered its re-routing code and so it was
7339 trying to route what was effectively the same domain again. This normally
7340 caused routing to fail (because the router wouldn't handle the domain
7341 twice). Now Exim checks for this case specially, and just changes the
7342 casing of the domain that it ultimately uses when it transmits the message
7345 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7348 50. If a filter generated a file delivery with a non-absolute name (possible if
7349 no home directory exists for the router), the forbid_file option was not
7352 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7353 the existing equality matching.
7355 52. Exim was using ints instead of ino_t variables in some places where it was
7356 dealing with inode numbers.
7358 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7359 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7360 called TMPDIR, and if it finds it is different, it changes its value.
7362 54. The smtp_printf() function is now made available to local_scan() so
7363 additional output lines can be written before returning. There is also an
7364 smtp_fflush() function to enable the detection of a dropped connection.
7365 The variables smtp_input and smtp_batched_input are exported to
7368 55. Changed the default runtime configuration: the message "Unknown user"
7369 has been removed from the ACL, and instead placed on the localuser router,
7370 using the cannot_route_message feature. This means that any verification
7371 failures that generate their own messages won't get overridden. Similarly,
7372 the "Unrouteable address" message that was in the ACL for unverifiable
7373 relay addresses has also been removed.
7375 56. Added hosts_avoid_esmtp to the smtp transport.
7377 57. The exicyclog script was not checking for the esoteric option
7378 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7379 will work only if exicyclog is run under the appropriate euid.
7381 58. Following a discussion on the list, the rules by which Exim recognises line
7382 endings on incoming messages have been changed. The -dropcr and drop_cr
7383 options are now no-ops, retained only for backwards compatibility. The
7384 following line terminators are recognized: LF CRLF CR. However, special
7385 processing applies to CR:
7387 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7388 nor a local message in the state where . is a terminator.
7390 (ii) If a bare CR is encountered in a header line, an extra space is added
7391 after the line terminator so as not to end the header. The reasoning
7392 behind this is that bare CRs in header lines are most likely either
7393 to be mistakes, or people trying to play silly games.
7395 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7396 was being incorrectly given as 18 bytes larger than it should have been.
7397 This is a VOB (very old bug).
7399 60. This may never have affected anything current, but just in case it has:
7400 When the local host is found other than at the start of a list of hosts,
7401 the local host, those with the same MX, and any that follow, are discarded.
7402 When the list in question was part of a longer list of hosts, the following
7403 hosts (not currently being processed) were also being discarded. This no
7404 longer happens. I'm not sure if this situation could ever has previously
7407 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7410 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7411 header. This is something that is recommended in a new Internet Draft, and
7412 is something that is documented as being done by Sendmail. There are two
7413 possible values. For messages generated by the autoreply transport, Exim
7416 Auto-Submitted: auto-replied
7418 whereas for all other generated messages (e.g. bounces) it adds
7420 Auto-Submitted: auto-generated
7422 63. The "personal" condition in filters now includes a test for the
7423 Auto-Submitted: header. If it contains the string "auto-" the message it
7424 not considered personal.
7426 64. Added rcpt_include_affixes as a generic transport option.
7428 65. Added queue_only_override (default true).
7430 66. Added the syslog_duplication option.
7432 67. If what should have been the first header line of a message consisted of
7433 a space followed by a colon, Exim was mis-interpreting it as a header line.
7434 It isn't of course - it is syntactically invalid and should therefore be
7435 treated as the start of the message body. The misbehaviour could have
7436 caused a number of strange effects, including loss of data in subsequent
7437 header lines, and spool format errors.
7439 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7440 client host had authenticated. This control can now be exercised by an ACL
7441 for more flexibility.
7443 69. By default, callouts do not happen when testing with -bh. There is now a
7444 variant, -bhc, which does actually run the callout code, including
7445 consulting and updating the callout cache.
7447 70. Added support for saslauthd authentication, courtesy of Alexander
7450 71. If statvfs() failed on the spool or log directories while checking their
7451 size for availability, Exim confusingly gave the error "space shortage".
7452 Furthermore, in debugging mode it crashed with a floating point exception.
7453 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7454 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7455 serious problem, Exim now writes to the main and panic logs when this
7456 happens, with details of the failure. It then refuses to accept the
7457 incoming message, giving the message "spool directory problem" or "log
7458 directory problem" with a 421 code for SMTP messages.
7460 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7461 0, 1, and 2 exist, because some OS complain for execs without them (see
7462 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7463 descriptors. However, the code omitted to check that the open succeeded,
7464 causing mysterious errors if for some reason the permissions on /dev/null
7465 got screwed. Now Exim writes a message to the main and panic logs, and
7466 bombs out if it can't open /dev/null.
7468 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7469 interact so that it is all more flexible. It is supposed to remain
7470 backwards compatible. Also added extra_local_interfaces.
7472 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7473 to bomb out with an assertion failure - to the client this appears as a
7474 connection drop. This problem occurs in the part of the code that was taken
7475 from the Samba project. Fortunately, the assertion is in a very simple
7476 function, so I have fixed this by reproducing the function inline in the
7477 one place where it is called, and arranging for authentication to fail
7478 instead of killing the process with assert().
7480 75. The SPA client code was not working when the server requested OEM rather
7481 than Unicode encoding.
7483 76. Added code to make require_files with a specific uid setting more usable in
7484 the case where statting the file as root fails - usually a non-root-mounted
7485 NFS file system. When this happens and the failure is EACCES, Exim now
7486 forks a subprocess and does the per-uid checking as the relevant uid.
7488 77. Added process_log_path.
7490 78. If log_file_path was not explicitly set, a setting of check_log_space or
7491 check_log_inodes was ignored.
7493 79. If a space check for the spool or log partitions fails, the incident is now
7494 logged. Of course, in the latter case the data may get lost...
7496 80. Added the %p formatting code to string_format() so that it can be used to
7497 print addresses in debug_print(). Adjusted all the address printing in the
7498 debugging in store.c to use %p rather than %d.
7500 81. There was a concern that a line of code in smtp_in.c could overflow a
7501 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7502 initially expressed, the concern was not well-founded, because trailing
7503 spaces are removed early. However, if the trailing spaces were followed by
7504 a NULL, they did not get removed, so the overflow was possible. Two fixes
7507 (a) I re-wrote the offending code in a cleaner fashion.
7508 (b) If an incoming SMTP command contains a NULL character, it is rejected
7511 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7512 runs initgroups(), so that if the Exim user is in any additional groups,
7513 they will be used during message reception.
7519 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob