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/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
12 buffer overrun for (non-chunking) other transports.
14 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
15 TLS1.3, means that a server rejecting a client certificate is not visible
16 to the client until the first read of encrypted data (typically the
17 response to EHLO). Add detection for that case and treat it as a failed
18 TLS connection attempt, so that the normal retry-in-clear can work (if
21 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
22 and/or domain. Found and fixed by Jason Betts.
24 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
25 configuration). If a CNAME target was not a wellformed name pattern, a
28 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
29 the OS reports them interleaved with other addresses.
31 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
32 used both for input and for a verify callout, both encrypted, SMTP
33 responses being sent by the server could be lost. This resulted in
34 dropped connections and sometimes bounces generated by a peer sending
37 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
38 string. Previously it was possible to cause undefined behaviour in a
39 library routine (usually a crash). Found by "zerons".
41 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
44 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
45 API was removed, so update to use the newer ones.
48 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
49 to changes apparently associated with TLS1.3 handling some of the APIs
50 previously used were either nonfunctional or inappropriate. Strings
51 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
52 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
53 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
54 This affects log line X= elements, the $tls_{in,out}_cipher variables,
55 and the use of specific cipher names in the encrypted= ACL condition.
57 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
58 verification result was not updated unless hosts_require_ocsp applied.
60 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
61 queue_list_requires_admin set to false, non-admin users were denied the
64 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
65 directory-of-certs mode. Previously they were advertised despite the
68 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
69 success-DSN messages. Previously the From: header was always the default
70 one for these; the option was ignored.
76 JH/01 Remove code calling the customisable local_scan function, unless a new
77 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
79 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
80 non-signal-safe functions being used.
82 JH/03 Bug 2269: When presented with a received message having a stupidly large
83 number of DKIM-Signature headers, disable DKIM verification to avoid
84 a resource-consumption attack. The limit is set at twenty.
86 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
87 report of oldest_pass in ${authres } in consequence, and separate out
88 some descriptions of reasons for verification fail.
90 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
91 files in the spool were present and unlocked. A queue-runner could spot
92 them, resulting in a duplicate delivery. Fix that by doing the unlock
93 after the unlink. Investigation by Tim Stewart. Take the opportunity to
94 add more error-checking on spoolfile handling while that code is being
97 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
98 No known attacks, no CVE, this is defensive hardening.
100 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
101 a queue-runner could start a delivery while other operations were ongoing.
102 Cutthrough delivery was a common victim, resulting in duplicate delivery.
103 Found and investigated by Tim Stewart. Fix by using the open message data
104 file handle rather than opening another, and not locally closing it (which
105 releases a lock) for that case, while creating the temporary .eml format
106 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
108 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
109 $sender_verify_failure/$recipient_verify_failure to "random".
111 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
114 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
115 Previously this would segfault.
117 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
120 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
121 like zero, since the resolver should be doing this for us, But we need one
122 as a CNAME but no MX presence gets the CNAME returned; we need to check
123 that doesn't point to an MX to declare it "no MX returned" rather than
124 "error, loop". A new main option is added so the older capability of
125 following some limited number of chain links is maintained.
127 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
129 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
130 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
131 now-deprecated earlier definitions used only specified the range up to TLS
132 1.2 (in the older-version library docs).
134 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
136 JH/15 Rework TLS client-side context management. Stop using a global, and
137 explicitly pass a context around. This enables future use of TLS for
138 connections to service-daemons (eg. malware scanning) while a client smtp
139 connection is using TLS; with cutthrough connections this is quite likely.
141 JH/16 Fix ARC verification to do AS checks in reverse order.
143 JH/17 Support a "tls" option on the ${readsocket } expansion item.
145 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
146 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
147 Previously the "utf8" would be re-prepended for every additional message.
149 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
150 Previously thery were accepted, resulting in issues when attempting to
151 forward messages to a non-supporting MTA.
153 PP/02 Let -n work with printing macros too, not just options.
155 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
156 one parent address was copied, and bogus data was used at delivery-logging
157 time. Either a crash (after delivery) or bogus log data could result.
158 Discovery and analysis by Tim Stewart.
160 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
161 Previously if the string ended mid-character, we did not insert the
162 promised '?' replacement.
164 PP/04 Documentation: current string operators work on bytes, not codepoints.
166 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
167 should pack well giving a smaller memory footprint so better caching and
168 therefore performance. Group the declarations where this can't be done so
169 that the byte-sized flag variables are not interspersed among pointer
170 variables, giving a better chance of good packing by the compiler.
172 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
173 non-null, to avoid issues with sites running BATV. Previously reports were
174 sent with an empty envelope sender so looked like bounces.
176 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
177 The ignore_error flag wasn't being returned from the filter subprocess so
178 was not set for later routers. Investigation and fix by Matthias Kurz.
180 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
181 and a msg:complete for the whole, when a message is manually removed using
182 -Mrm. Developement by Matthias Kurz, hacked on by JH.
184 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
185 a "Gnu special" function, asprintf() in the DB utility binary builds; I
186 hope that is portable enough.
188 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
189 requiring a known-CA anchor certificate; make it now rely entirely on the
190 TLSA as an anchor. Checking the name on the leaf cert against the name
191 on the A-record for the host is still done for TA (but not for EE mode).
193 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
194 included in delivery lines for non-proxied connections, when compiled with
195 SUPPORT_SOCKS and running with proxy logging enabled.
197 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
198 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
199 move the existing event to fire before the normal logging of message
200 failure so that custom logging is bracketed by normal logging.
202 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
203 msg:fail:internal event. Developement by Matthias Kurz.
205 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
206 far too small for todays use of crypto signatures stored there. Go all
207 the way to the max DNS message size of 64kB, even though this might be
208 overmuch for IOT constrained device use.
210 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
211 copy a string over itself. The library routine is documented as not
212 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
214 JH/32 For main options check_spool_space and check_inode_space, where the
215 platform supports 64b integers, support more than the previous 2^31 kB
216 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
217 the previous G, M, k.
219 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
220 $authenticated_fail_id variable on authentication failure. Previously
223 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
224 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
225 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
226 GNUTLS_SEC_PARAM_MEDIUM.
228 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
229 side. Previously we would continue as if no SNI had been received.
231 JH/36 Harden the handling of string-lists. When a list consisted of a sole
232 "<" character, which should be a list-separator specification, we walked
233 off past the nul-terimation.
235 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
236 causes) even when the retry time is not yet met. Previously they were
237 not, meaning that when (say) an account was over-quota and temp-rejecting,
238 and multiple senders' messages were queued, only one sender would get
239 notified on each configured delay_warning cycle.
241 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
243 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
244 cipher-suites, an error can be left on the stack even for a succeeding
245 accept; this results in impossible error messages when a later operation
248 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
249 return error codes indicating retry. Under TLS1.3 this becomes required.
251 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
252 it only wrote the new authenticators, resulting in a lack of tracking of
253 peer changes of ESMTP extensions until the next cache flush.
255 JH/41 Fix the loop reading a message header line to check for integer overflow,
256 and more-often against header_maxsize. Previously a crafted message could
257 induce a crash of the recive process; now the message is cleanly rejected.
259 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
260 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
266 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
267 When redis_servers is set to a list of > 1 element, and the Redis servers
268 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
269 case of MOVED into a DEFER case instead, thus moving the query onto the
270 next server in the list. For a cluster of N elements, all N servers must
271 be defined in redis_servers.
273 GF/02 Catch and remove uninitialized value warning in exiqsumm
274 Check for existence of @ARGV before looking at $ARGV[0]
276 JH/01 Replace the store_release() internal interface with store_newblock(),
277 which internalises the check required to safely use the old one, plus
278 the allocate and data copy operations duplicated in both (!) of the
279 extant use locations.
281 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
282 modifier. This matches the restriction on the commandline.
284 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
285 Previously only the last row was returned.
287 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
288 we assumed that tags in the header were well-formed, and parsed the
289 element content after inspecting only the first char of the tag.
290 Assumptions at that stage could crash the receive process on malformed
293 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
294 While running the DKIM ACL we operate on the Permanent memory pool so that
295 variables created with "set" persist to the DATA ACL. Also (at any time)
296 DNS lookups that fail create cache records using the Permanent pool. But
297 expansions release any allocations made on the current pool - so a dnsdb
298 lookup expansion done in the DKIM ACL releases the memory used for the
299 DNS negative-cache, and bad things result. Solution is to switch to the
300 Main pool for expansions.
301 While we're in that code, add checks on the DNS cache during store_reset,
302 active in the testsuite.
303 Problem spotted, and debugging aided, by Wolfgang Breyha.
305 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
306 When none of the hosts presented to a transport match an already-open
307 connection, close it and proceed with the list. Previously we would
308 queue the message. Spotted by Lena with Yahoo, probably involving
311 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
312 Previously a spurious "250 OK id=" response was appended to the proper
315 JH/08 The "support for" informational output now, which built with Content
316 Scanning support, has a line for the malware scanner interfaces compiled
317 in. Interface can be individually included or not at build time.
319 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
320 by the template makefile "src/EDITME". The "STREAM" support for an older
321 ClamAV interface method is removed.
323 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
324 rows affected is given instead).
326 JH/11 The runtime Berkeley DB library version is now additionally output by
327 "exim -d -bV". Previously only the compile-time version was shown.
329 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
330 SMTP connection. Previously, when one had more receipients than the
331 first, an abortive onward connection was made. Move to full support for
332 multiple onward connections in sequence, handling cutthrough connection
333 for all multi-message initiating connections.
335 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
336 routers. Previously, a multi-recipient message would fail to match the
337 onward-connection opened for the first recipient, and cause its closure.
339 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
340 a timeout on read on a GnuTLS initiating connection, resulting in the
341 initiating connection being dropped. This mattered most when the callout
342 was marked defer_ok. Fix to keep the two timeout-detection methods
345 JH/15 Relax results from ACL control request to enable cutthrough, in
346 unsupported situations, from error to silently (except under debug)
347 ignoring. This covers use with PRDR, frozen messages, queue-only and
350 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
352 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
353 metadata, resulting in a crash in free().
355 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
356 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
357 Broken also in d185889f4, with init system revamp.
359 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
360 Previously we abruptly closed the connection after reading a malware-
361 found indication; now we go on to read the "scan ok" response line,
364 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
365 ACL. Previously, a crash would result.
367 JH/19 Speed up macro lookups during configuration file read, by skipping non-
368 macro text after a replacement (previously it was only once per line) and
369 by skipping builtin macros when searching for an uppercase lead character.
371 JH/20 DANE support moved from Experimental to mainline. The Makefile control
372 for the build is renamed.
374 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
375 was allocated for every new TLS startup, meaning one per message. Fix
376 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
378 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
379 reported the original. Fix to report (as far as possible) the ACL
380 result replacing the original.
382 JH/23 Fix memory leak during multi-message connections using STARTTLS under
383 OpenSSL. Certificate information is loaded for every new TLS startup,
384 and the resources needed to be freed.
386 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
388 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
391 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
392 DATA response info to the (existing) per-recipient response info for
393 the "C=" log element. It can have useful tracking info from the
394 destination system. Patch from Simon Arlott.
396 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
397 length value. Previously this would segfault.
399 HS/02 Support Avast multiline protoocol, this allows passing flags to
400 newer versions of the scanner.
402 JH/28 Ensure that variables possibly set during message acceptance are marked
403 dead before release of memory in the daemon loop. This stops complaints
404 about them when the debug_store option is enabled. Discovered specifically
405 for sender_rate_period, but applies to a whole set of variables.
406 Do the same for the queue-runner and queue-list loops, for variables set
407 from spool message files. Do the same for the SMTP per-message loop, for
408 certain variables indirectly set in ACL operations.
410 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
411 as a multi-recipient message from a mailinglist manager). The coding had
412 an arbitrary cutoff number of characters while checking for more input;
413 enforced by writing a NUL into the buffer. This corrupted long / fast
414 input. The problem was exposed more widely when more pipelineing of SMTP
415 responses was introduced, and one Exim system was feeding another.
416 The symptom is log complaints of SMTP syntax error (NUL chars) on the
417 receiving system, and refused recipients seen by the sending system
418 (propating to people being dropped from mailing lists).
419 Discovered and pinpointed by David Carter.
421 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
422 replaced by the ${authresults } expansion.
424 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
426 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
427 allows proper process termination in container environments.
429 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
430 Previously the "final dot" had a newline after it; ensure it is CR,LF.
432 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
433 and "err_perm", deprecated since 4.83 when the RFC-defined words
434 "temperror" and "permerror" were introduced.
436 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
437 transport-filters or DKIM-signing. The restriction was lost in the
438 consolidation of verify-callout and delivery SMTP handling.
439 Extend the restriction to also cover ARC-signing.
441 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
442 in defer=pass mode supply a 450 to the initiator. Previously the message
445 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
446 tls_require_ciphers is used as before.
448 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
449 "pass_unscanned". Only tmpfails from the scanner are written to
450 the paniclog, as they may require admin intervention (permission
451 denied, license issues). Other scanner errors (like decompression
452 bombs) do not cause a paniclog entry.
454 JH/36 Fix reinitialisation of DKIM logging variable between messages.
455 Previously it was possible to log spurious information in receive log
458 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
459 triggered odd behaviour from Outlook Express clients.
461 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
464 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
465 since the IETF WG has not yet settled on that versus the original
466 "bare" representation.
468 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
469 Previously the millisecond value corrupted the output.
470 Fix also for syslog_pid=no and log_selector +pid, for which the pid
471 corrupted the output.
477 JH/01 Rework error string handling in TLS interface so that the caller in
478 more cases is responsible for logging. This permits library-sourced
479 string to be attached to addresses during delivery, and collapses
480 pairs of long lines into single ones.
482 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
483 during configuration. Wildcards are allowed and expanded.
485 JH/02 Rework error string handling in DKIM to pass more info back to callers.
486 This permits better logging.
488 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
489 do not close it down and have the child transport start it up again on
490 the passed-on TCP connection. Instead, proxy the child (and any
491 subsequent ones) for TLS via a unix-domain socket channel. Logging is
492 affected: the continued delivery log lines do not have any DNSSEC, TLS
493 Certificate or OCSP information. TLS cipher information is still logged.
495 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
496 identical IP addresses on different listening ports. Will also affect
499 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
500 add noisy ifdef guards to special-case this sillyness.
501 Patch from Bernd Kuhls.
503 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
504 than 255 are no longer allowed.
506 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
507 Disable the session-cache too, which might reduce our load. Since we
508 currrectly use a new context for every connection, both as server and
509 client, there is no benefit for these.
510 GnuTLS appears to not support tickets server-side by default (we don't
511 call gnutls_session_ticket_enable_server()) but client side is enabled
512 by default on recent versions (3.1.3 +) unless the PFS priority string
515 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
516 <https://reproducible-builds.org/specs/source-date-epoch/>.
518 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
519 the check for any unsuccessful recipients did not notice the limit, and
520 erroneously found still-pending ones.
522 JH/08 Pipeline CHUNKING command and data together, on kernels that support
523 MSG_MORE. Only in-clear (not on TLS connections).
525 JH/09 Avoid using a temporary file during transport using dkim. Unless a
526 transport-filter is involved we can buffer the headers in memory for
527 creating the signature, and read the spool data file once for the
528 signature and again for transmission.
530 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
531 4.77 as the kernel support then wasn't solid, having issues in 64bit
532 mode. Now, it's been long enough. Add support for FreeBSD also.
534 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
535 case where the routing stage had gathered several addresses to send to
536 a host before calling the transport for the first, we previously failed
537 to close down TLS in the old transport process before passing the TCP
538 connection to the new process. The new one sent a STARTTLS command
539 which naturally failed, giving a failed delivery and bloating the retry
540 database. Investigation and fix prototype from Wolfgang Breyha.
542 JH/12 Fix check on SMTP command input synchronisation. Previously there were
543 false-negatives in the check that the sender had not preempted a response
544 or prompt from Exim (running as a server), due to that code's lack of
545 awareness of the SMTP input buffering.
547 PP/04 Add commandline_checks_require_admin option.
548 Exim drops privileges sanely, various checks such as -be aren't a
549 security problem, as long as you trust local users with access to their
550 own account. When invoked by services which pass untrusted data to
551 Exim, this might be an issue. Set this option in main configuration
552 AND make fixes to the calling application, such as using `--` to stop
555 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
556 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
557 responses to those, into a single TLS record each way (this usually means
558 a single packet). As a side issue, smtp_enforce_sync now works on TLS
561 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
562 affects you only if you're dancing at the edge of the param size limits.
563 If you are, and this message makes sense to you, then: raise the
564 configured limit or use OpenSSL 1.1. Nothing we can do for older
567 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
568 cmdline element to get the documented default one. Previously it was
571 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
572 are unsuable for non-message-specific reasons - eg. connection timeout,
573 banner-time rejection.
575 JH/16 Fix logging of delivery remote port, when specified by router, under
578 PP/06 Repair manualroute's ability to take options in any order, even if one
579 is the name of a transport.
582 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
584 JH/17 Change the list-building routines interface to use the expanding-string
585 triplet model, for better allocation and copying behaviour.
587 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
588 Previously it was constructed the first time a possibly-matching string
589 was met in the configuration file input during startup; now it is done
592 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
593 compatible one, to avoid the (poorly documented) possibility of a config
594 file in the working directory redirecting the DB files, possibly correpting
595 some existing file. CVE-2017-10140 assigned for BDB.
597 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
598 cache-hot. Previously, although the result was properly cached, the
599 initial verify call returned a defer.
601 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
602 the main verify for receipient in uncached-mode.
604 JH/22 Retire historical build files to an "unsupported" subdir. These are
605 defined as "ones for which we have no current evidence of testing".
607 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
608 if present. Previously it was ignored.
610 JH/24 Start using specified-initialisers in C structure init coding. This is
611 a C99 feature (it's 2017, so now considered safe).
613 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
614 if was a fixed-sized field and bitmask ops via macros; it is now more
617 PP/07 GitHub PR 56: Apply MariaDB build fix.
618 Patch provided by Jaroslav Škarvada.
620 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
621 during Coverity cleanups [4.87 JH/47]
622 Diagnosis and fix provided by Michael Fischer v. Mollard.
624 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
625 the right size to place the terminating semicolon on its own folded
626 line, the header hash was calculated to an incorrect value thanks to
627 the (relaxed) space the fold became.
629 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
630 and confused the parent.
632 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
633 which could crash as a result. This could lead to undeliverable messages.
635 JH/28 Logging: "next input sent too soon" now shows where input was truncated
638 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
639 matters on fast-turnover and PID-randomising systems, which were getting
640 out-of-order delivery.
642 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
643 a possibly-overlapping copy. The symptom was that "Remote host closed
644 connection in response to HELO" was logged instead of the actual 4xx
647 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
648 Previously only that bufferd was discarded, resulting in SYMTP command
651 JH/32 DKIM: when a message has multiple signatures matching an identity given
652 in dkim_verify_signers, run the dkim acl once for each. Previously only
653 one run was done. Bug 2189.
655 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
656 "panic the current process" to "deliberately defer". The panic log is
657 still written with the problem list name; the mail and reject logs now
658 get a temp-reject line for the message that was being handled, saying
659 something like "domains check lookup or other defer". The SMTP 451
660 message is still "Temporary local problem".
662 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
663 A crafted sequence of BDAT commands could result in in-use memory beeing
664 freed. CVE-2017-16943.
666 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
667 from SMTP input. Previously it was always done; now only done for DATA
668 and not BDAT commands. CVE-2017-16944.
670 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
671 to the message (such as an overlong header line). Previously this was
672 not done and we did not exit BDAT mode. Followon from the previous item
673 though a different problem.
679 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
680 than -2003 did; needs libidn2 in addition to libidn.
682 JH/02 The path option on a pipe transport is now expanded before use.
684 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
685 Patch provided by "Björn", documentation fix added too.
687 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
688 missing a wire-to-host endian conversion.
690 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
691 close after a BDAT command line could be taken as a following command,
692 giving a synch failure. Fix by only checking for synch immediately
693 before acknowledging the chunk.
695 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
696 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
697 macro. Patches provided by Josh Soref.
699 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
700 Previously we did not; the RFC seems ambiguous and VRFY is not listed
701 by IANA as a service extension. However, John Klensin suggests that we
704 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
705 the dkim code may be unix-mode line endings rather than smtp wire-format
706 CRLF, so prepend a CR to any bare LF.
708 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
709 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
711 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
712 the first were themselves being wrongly included in the feed into dkim
713 processing; with most chunk sizes in use this resulted in an incorrect
714 body hash calculated value.
716 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
717 DKIM signature block, for verification. Although advised against by
718 standards it is specifically not ruled illegal.
720 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
722 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
723 missing a body hash (the bh= tag).
725 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
726 It seems that HAProxy sends the Proxy Protocol information in clear and
727 only then does a TLS startup, so do the same.
729 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
730 TCP connections (such as for Spamd) unless the daemon successfully set
731 Fast Open mode on its listening sockets. This fixes breakage seen on
732 too-old kernels or those not configured for Fast Open, at the cost of
733 requiring both directions being enabled for TFO, and TFO never being used
734 by non-daemon-related Exim processes.
736 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
737 endings, at least on the first header line. Try to canonify any that get
738 past that check, despite the cost.
740 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
741 now limited to an arbitrary five deep, while parsing addresses with the
742 strip_excess_angle_brackets option enabled.
744 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
745 instead leave the unprompted TLS handshake in socket buffer for the
746 TLS library to consume.
748 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
750 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
752 JH/16 Drop variables when they go out of scope. Memory management drops a whole
753 region in one operation, for speed, and this leaves assigned pointers
754 dangling. Add checks run only under the testsuite which checks all
755 variables at a store-reset and panics on a dangling pointer; add code
756 explicitly nulling out all the variables discovered. Fixes one known
757 bug: a transport crash, where a dangling pointer for $sending_ip_address
758 originally assigned in a verify callout, is re-used.
760 PP/06 Drop '.' from @INC in various Perl scripts.
762 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
764 PP/08 Reduce a number of compilation warnings under clang; building with
765 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
766 should be warning-free.
768 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
770 HS/01 Fix portability problems introduced by PP/08 for platforms where
771 realloc(NULL) is not equivalent to malloc() [SunOS et al].
773 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
774 chunk. This allows us to accept broken chunked messages. We need a more
775 general solution here.
777 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
778 already-broken messages in the queue.
780 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
782 JH/19 Fix reference counting bug in routing-generated-address tracking.
788 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
789 supports it and a size is available (ie. the sending peer gave us one).
791 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
792 years of being deprecated). The replacements are the ACLs
793 acl_smtp_mime and acl_not_smtp_mime.
795 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
796 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
797 or even an in-clear connection were permitted. Now, if the host lookup
798 was dnssec and dane was requested then the host is only used if the
799 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
800 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
801 if one fails this test.
802 This means that a poorly-configured remote DNS will make it incommunicado;
803 but it protects against a DNS-interception attack on it.
805 JH/04 Bug 1810: make continued-use of an open smtp transport connection
806 non-noisy when a race steals the message being considered.
808 JH/05 If main configuration option tls_certificate is unset, generate a
809 self-signed certificate for inbound TLS connections.
811 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
812 in rewrites and routers.
814 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
815 and logged a warning sing 4.83; now they are a configuration file error.
817 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
818 (lacking @domain). Apply the same qualification processing as RCPT.
820 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
822 JH/10 Support ${sha256:} applied to a string (as well as the previous
825 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
826 a cutthrough deliver is pending, as we always want to make a connection.
827 This also avoids re-routing the message when later placing the cutthrough
828 connection after a verify cache hit.
829 Do not update it with the verify result either.
831 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
832 when routing results in more than one destination address.
834 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
835 signing (which inhibits the cutthrough capability). Previously only
836 the presence of an option was tested; now an expansion evaluating as
837 empty is permissible (obviously it should depend only on data available
838 when the cutthrough connection is made).
840 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
841 the relevant preceding SMTP command did not note the pipelining mode.
843 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
844 Previously they were not counted.
846 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
847 as one having no matching records. Previously we deferred the message
848 that needed the lookup.
850 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
851 distinguished as "(=".
853 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
854 for missing MX records. Previously it only worked for missing A records.
856 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
858 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
859 after the data-go-ahead and data-ack. Patch from Jason Betts.
861 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
862 even for a "none" policy. Patch from Tony Meyer.
864 JH/22 Fix continued use of a connection for further deliveries. If a port was
865 specified by a router, it must also match for the delivery to be
868 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
869 When one of the recipients of a message was unsuitable for the connection
870 (has no matching addresses), we lost track of needing to mark it
871 deferred. As a result mail would be lost.
873 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
875 JH/25 Decoding ACL controls is now done using a binary search; the source code
876 takes up less space and should be simpler to maintain. Merge the ACL
877 condition decode tables also, with similar effect.
879 JH/26 Fix problem with one_time used on a redirect router which returned the
880 parent address unchanged. A retry would see the parent address marked as
881 delivered, so not attempt the (identical) child. As a result mail would
884 JH/27 Fix a possible security hole, wherein a process operating with the Exim
885 UID can gain a root shell. Credit to http://www.halfdog.net/ for
886 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
889 JH/28 Enable {spool,log} filesystem space and inode checks as default.
890 Main config options check_{log,spool}_{inodes,space} are now
891 100 inodes, 10MB unless set otherwise in the configuration.
893 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
894 Previously it only applied to the main-section connection policy
897 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
899 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
900 by me. Added RFC7919 DH primes as an alternative.
902 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
903 are not in the system include path.
905 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
906 GnuTLS, when a session startup failed (eg because the client disconnected)
907 Exim did stdio operations after fclose. This was exposed by a recent
908 change which nulled out the file handle after the fclose.
910 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
911 signed directly by the cert-signing cert, rather than an intermediate
912 OCSP-signing cert. This is the model used by LetsEncrypt.
914 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
916 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
917 an incoming connection.
919 HS/02 Bug 1802: Do not half-close the connection after sending a request
922 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
923 fallback to "prime256v1".
925 JH/34 SECURITY: Use proper copy of DATA command in error message.
926 Could leak key material. Remotely exploitable. CVE-2016-9963.
932 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
933 and 3.4.4 - once the server is enabled to respond to an OCSP request
934 it does even when not requested, resulting in a stapling non-aware
935 client dropping the TLS connection.
937 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
938 support variable-length bit vectors. No functional change.
940 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
941 The I= interface field on outgoing lines is now after the H= remote
942 host field, same as incoming lines. There is a separate
943 outgoing_interface log selector which allows you to disable the
946 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
947 If not running log_selector +smtp_connection the mainlog would be held
948 open indefinitely after a "too many connections" event, including to a
949 deleted file after a log rotate. Leave the per net connection logging
950 leaving it open for efficiency as that will be quickly detected by the
951 check on the next write.
953 HS/01 Bug 1671: Fix post transport crash.
954 Processing the wait-<transport> messages could crash the delivery
955 process if the message IDs didn't exist for some reason. When
956 using 'split_spool_directory=yes' the construction of the spool
957 file name failed already, exposing the same netto behaviour.
959 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
960 mime_regex ACL conditions.
962 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
963 to DSN fail messages (bounces): remote IP, remote greeting, remote response
964 to HELO, local diagnostic string.
966 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
967 log line to debug. Even when configured with a tls authenticator many
968 client connections are expected to not authenticate in this way, so
969 an authenticate fail is not an error.
971 HS/02 Add the Exim version string to the process info. This way exiwhat
972 gives some more detail about the running daemon.
974 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
975 matter for fast-change records such as DNSBLs.
977 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
978 retry record, even if constant. There may be multiple transports with
979 different interface settings and the retry behaviour needs to be kept
982 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
984 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
986 JH/10 Bug 840: fix log_defer_output option of pipe transport
988 JH/11 Bug 830: use same host for all RCPTS of a message, even under
989 hosts_randomize. This matters a lot when combined with mua_wrapper.
991 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
992 ${quote_pgsql:<string>} operator.
994 JH/13 Bug 1708: avoid misaligned access in cached lookup.
996 JH/14 Change header file name for freeradius-client. Relevant if compiling
997 with Radius support; from the Gentoo tree and checked under Fedora.
999 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1001 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1002 option transport_filter, meaning no filtering.
1004 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1006 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1007 defaults to "*" (all hosts). The variable is now available when not built
1008 with TLS, default unset, mainly to enable keeping the testsuite sane.
1009 If a server certificate is not supplied (via tls_certificate) an error is
1010 logged, and clients will find TLS connections fail on startup. Presumably
1011 they will retry in-clear.
1012 Packagers of Exim are strongly encouraged to create a server certificate
1013 at installation time.
1015 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1016 with the $config_file variable.
1018 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1019 in transport context, after the attempt, and per-recipient. The latter type
1020 is per host attempted. The event data is the error message, and the errno
1021 information encodes the lookup type (A vs. MX) used for the (first) host,
1022 and the trailing two digits of the smtp 4xx response.
1024 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1025 to write to mainlog (or rejectlog, paniclog) in the window between file
1026 creation and permissions/ownership being changed. Particularly affects
1027 installations where exicyclog is run as root, rather than exim user;
1028 result is that the running daemon panics and dies.
1030 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1032 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1033 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1034 "pri" and "weight". Note that the previous implicit priority given by the
1035 list order is no longer honoured.
1037 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1038 for DKIM processing.
1040 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1041 by defining SUPPORT_SOCKS.
1043 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1044 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1045 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1046 variables are renamed to proxy_{local,external}_{address,port}.
1048 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1049 for a build by defining SUPPORT_I18N
1051 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1052 of the query string, and make ${quote_redis:} do that quoting.
1054 JH/29 Move Events support from Experimental to mainline, enabled by default
1055 and removable for a build by defining DISABLE_EVENT.
1057 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1059 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1060 cached by the daemon.
1062 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1063 by defining LOOKUP_REDIS. The libhiredis library is required.
1065 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1066 keys are given for lookup.
1068 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1069 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1070 only supported when built with TLS support. The PolarSSL SHA routines
1071 are still used when the TLS library is too old for convenient support.
1073 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1074 openssl_options), for security. OpenSSL forces this from version 1.1.0
1075 server-side so match that on older versions.
1077 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1078 allocation for $value could be released as the expansion processing
1079 concluded, but leaving the global pointer active for it.
1081 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1082 and to use the domains and local_parts ACL conditions.
1084 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1085 incorrectly not doubled on cutthrough transmission, hence seen as a
1086 body-termination at the receiving system - resulting in truncated mails.
1087 Commonly the sender saw a TCP-level error, and retransmitted the message
1088 via the normal store-and-forward channel. This could result in duplicates
1089 received - but deduplicating mailstores were liable to retain only the
1090 initial truncated version.
1092 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1094 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1096 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1097 we're in there, support oversigning also; bug 1309.
1099 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1101 HS/04 Add support for keep_environment and add_environment options.
1103 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1104 either intentional arithmetic overflow during PRNG, or testing config-
1107 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1108 delivery resulted in actual delivery. Cancel cutthrough before DATA
1111 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1112 rejects a recipient: pass the reject to the originator.
1114 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1115 Many were false-positives and ignorable, but it's worth fixing the
1118 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1119 for the new environment-manipulation done at startup. Move the routines
1120 from being local to tls.c to being global via the os.c file.
1122 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1123 an extract embedded as result-arg for a map, the first arg for extract
1124 is unavailable so we cannot tell if this is a numbered or keyed
1125 extraction. Accept either.
1131 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1134 JH/02 The smtp transport option "multi_domain" is now expanded.
1136 JH/03 The smtp transport now requests PRDR by default, if the server offers
1139 JH/04 Certificate name checking on server certificates, when exim is a client,
1140 is now done by default. The transport option tls_verify_cert_hostnames
1141 can be used to disable this per-host. The build option
1142 EXPERIMENTAL_CERTNAMES is withdrawn.
1144 JH/05 The value of the tls_verify_certificates smtp transport and main options
1145 default to the word "system" to access the system default CA bundle.
1146 For GnuTLS, only version 3.0.20 or later.
1148 JH/06 Verification of the server certificate for a TLS connection is now tried
1149 (but not required) by default. The verification status is now logged by
1150 default, for both outbound TLS and client-certificate supplying inbound
1153 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1156 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1157 Status Notification (bounce) messages are now MIME format per RFC 3464.
1158 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1159 under the control of the dsn_advertise_hosts option, and routers may
1160 have a dsn_lasthop option.
1162 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1163 default, modifiable by a malware= option. The list separator for
1164 the options can now be changed in the usual way. Bug 68.
1166 JH/10 The smtp_receive_timeout main option is now expanded before use.
1168 JH/11 The incoming_interface log option now also enables logging of the
1169 local interface on delivery outgoing connections.
1171 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1172 if the interface and destination host and port all match.
1174 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1177 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1178 Patch from Andrew Lewis.
1180 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1181 now supports optional time-restrictions, weighting, and priority
1182 modifiers per server. Patch originally by <rommer@active.by>.
1184 JH/16 The spamd_address main option now supports a mixed list of local
1185 and remote servers. Remote servers can be IPv6 addresses, and
1186 specify a port-range.
1188 JH/17 Bug 68: The spamd_address main option now supports an optional
1189 timeout value per server.
1191 JH/18 Bug 1581: Router and transport options headers_add/remove can
1192 now have the list separator specified.
1194 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1197 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1200 JH/21 Support for the A6 type of dns record is withdrawn.
1202 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1203 rather than the verbs used.
1205 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1206 from 255 to 1024 chars.
1208 JH/24 Verification callouts now attempt to use TLS by default.
1210 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1211 are generic router options now. The defaults didn't change.
1213 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1214 Original patch from Alexander Shikoff, worked over by JH.
1216 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1217 files if ZCAT_COMMAND is not executable.
1219 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1221 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1223 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1224 Normally benign, it bites when the pair was led to by a CNAME;
1225 modern usage is to not canonicalize the domain to a CNAME target
1226 (and we were inconsistent anyway for A-only vs AAAA+A).
1228 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1230 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1231 when evaluating $sender_host_dnssec.
1233 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1234 $sender_helo_dnssec variable.
1236 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1238 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1240 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1242 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1243 documented as working, but never had. Support all but $spam_report.
1245 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1246 added for tls authenticator.
1248 HS/03 Add perl_taintmode main config option
1254 TL/01 When running the test suite, the README says that variables such as
1255 no_msglog_check are global and can be placed anywhere in a specific
1256 test's script, however it was observed that placement needed to be near
1257 the beginning for it to behave that way. Changed the runtest perl
1258 script to read through the entire script once to detect and set these
1259 variables, reset to the beginning of the script, and then run through
1260 the script parsing/test process like normal.
1262 TL/02 The BSD's have an arc4random API. One of the functions to induce
1263 adding randomness was arc4random_stir(), but it has been removed in
1264 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1265 function when detected.
1267 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1268 cause callback expansion.
1270 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1271 syntax errors in an expansion can be treated as a string instead of
1272 logging or causing an error, due to the internal use of bool_lax
1273 instead of bool when processing it.
1275 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1276 server certificates when making smtp deliveries.
1278 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1280 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1282 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1284 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1285 Merged patch from Sebastian Wiedenroth.
1287 JH/05 Fix results-pipe from transport process. Several recipients, combined
1288 with certificate use, exposed issues where response data items split
1289 over buffer boundaries were not parsed properly. This eventually
1290 resulted in duplicates being sent. This issue only became common enough
1291 to notice due to the introduction of connection certificate information,
1292 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1294 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1295 size buffer was used, resulting in syntax errors when an expansion
1298 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1299 version 3.3.6 or later.
1301 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1302 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1303 both become "event_action", the variables become $event_name, $event_data
1304 and $event_defer_errno. There is a new variable $verify_mode, usable in
1305 routers, transports and related events. The tls:cert event is now also
1306 raised for inbound connections, if the main configuration event_action
1309 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1310 early OCSP support, but no stapling (appears to be less than 1.0.0).
1312 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1313 server certificate names available under the smtp transport option
1314 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1317 JH/10 Time-related extraction expansions from certificates now use the main
1318 option "timezone" setting for output formatting, and are consistent
1319 between OpenSSL and GnuTLS compilations. Bug 1541.
1321 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1322 encoded parameter in the incoming message. Bug 1558.
1324 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1325 include certificate info, eximon was claiming there were spoolfile
1328 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1330 JH/14 Log delivery-related information more consistently, using the sequence
1331 "H=<name> [<ip>]" wherever possible.
1333 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1334 are problematic for Debian distribution, omit them from the release
1337 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1339 JH/16 Fix string representation of time values on 64bit time_t architectures.
1342 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1343 output list separator was used.
1348 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1349 checkers that were complaining about end of non-void function with no
1352 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1353 This was a regression introduced in 4.83 by another bugfix.
1355 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1357 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1358 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1364 TF/01 Correctly close the server side of TLS when forking for delivery.
1366 When a message was received over SMTP with TLS, Exim failed to clear up
1367 the incoming connection properly after forking off the child process to
1368 deliver the message. In some situations the subsequent outgoing
1369 delivery connection happened to have the same fd number as the incoming
1370 connection previously had. Exim would try to use TLS and fail, logging
1371 a "Bad file descriptor" error.
1373 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1374 utilities have not been installed.
1376 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1377 temporary space as the ACL may create new global variables.
1379 TL/01 LDAP support uses per connection or global context settings, depending
1380 upon the detected version of the libraries at build time.
1382 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1383 to extract and use the src ip:port in logging and expansions as if it
1384 were a direct connection from the outside internet. PPv2 support was
1385 updated based on HAProxy spec change in May 2014.
1387 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1389 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1390 Properly escape header and check for NULL return.
1392 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1395 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1397 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1398 characters in header names, implemented as a verify condition.
1399 Contributed by Michael Fischer v. Mollard.
1401 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1402 results permerror and temperror. Previous values are deprecated but
1403 still accepted. In a future release, err_perm and err_temp will be
1404 completely removed, which will be a backward incompatibility if the
1405 ACL tests for either of these two old results. Patch contributed by
1406 user bes-internal on the mailing list.
1408 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1410 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1411 selectors, in both main and reject logs.
1413 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1416 JH/07 Add malware type "sock" for talking to simple daemon.
1418 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1420 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1421 routers/transports under cutthrough routing.
1423 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1424 numbers. Touch up "bool" conditional to keep the same definition.
1426 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1428 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1430 JH/12 Expand items in router/transport headers_add or headers_remove lists
1431 individually rather than the list as a whole. Bug 1452.
1433 Required for reasonable handling of multiple headers_ options when
1434 they may be empty; requires that headers_remove items with embedded
1435 colons must have them doubled (or the list-separator changed).
1437 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1438 view the policy declared in the DMARC record. Currently, $dmarc_status
1439 is a combined value of both the record presence and the result of the
1442 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1444 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1445 dnslookup router and the smtp transport (applying to the forward
1448 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1449 of ldap servers used for a specific lookup. Patch provided by Heiko
1452 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1453 New variable $lookup_dnssec_authenticated for observability.
1455 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1456 Patch submitted by Lars Timman.
1458 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1460 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1461 Requires trusted mode and valid format message id, aborts otherwise.
1462 Patch contributed by Heiko Schlichting.
1464 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1465 certextract with support for various fields. Bug 1358.
1467 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1468 is requested by default, modifiable by smtp transport option
1471 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1472 operate on certificate variables to give certificate fingerprints
1473 Also new ${sha256:cert_variable}.
1475 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1477 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1478 Christian Aistleitner.
1480 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1482 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1483 file. Patch from Wolfgang Breyha.
1485 JH/25 Expand the coverage of the delivery $host and $host_address to
1486 client authenticators run in verify callout. Bug 1476.
1488 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1489 align with daemon_smtp_ports. Bug 72.
1491 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1492 support and error reporting did not work properly.
1494 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1495 and is readable. Patch from Andrew Colin Kissa.
1497 TL/14 Enhance documentation of ${run expansion and how it parses the
1498 commandline after expansion, particularly in the case when an
1499 unquoted variable expansion results in an empty value.
1501 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1503 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1504 by renaming away from it. Addresses GH issue 15, reported by
1507 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1509 TL/15 SECURITY: prevent double expansion in math comparison functions
1510 (can expand unsanitized data). Not remotely exploitable.
1517 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1519 PP/02 Make -n do something, by making it not do something.
1520 When combined with -bP, the name of an option is not output.
1522 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1525 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1526 $sender_host_name and config options to manage this, and basic check
1529 PP/05 DSCP support for outbound connections and control modifier for inbound.
1531 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1532 (Only plugin which currently uses this is kerberos4, which nobody should
1533 be using, but we should make it available and other future plugins might
1534 conceivably use it, even though it would break NAT; stuff *should* be
1535 using channel bindings instead).
1537 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1538 name; added for Sendmail compatibility; requires admin caller.
1539 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1540 just ignore it); requires trusted caller.
1541 Also parse but ignore: -Ac -Am -X<logfile>
1544 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1546 TL/02 Add +smtp_confirmation as a default logging option.
1548 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1549 Patch by Magnus Holmgren from 2007-02-20.
1551 TL/04 Bugzilla 1281 - Spec typo.
1552 Bugzilla 1283 - Spec typo.
1553 Bugzilla 1290 - Spec grammar fixes.
1555 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1557 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1559 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1560 mailing list by by Dmitry Isaikin.
1562 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1564 JH/02 Support "G" suffix to numbers in ${if comparisons.
1566 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1568 NM/01 Bugzilla 1197 - Spec typo
1569 Bugzilla 1196 - Spec examples corrections
1571 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1573 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1574 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1577 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1578 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1580 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1581 "acl {{name}{arg}...}", and optional args on acl condition
1584 JH/05 Permit multiple router/transport headers_add/remove lines.
1586 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1588 JH/07 Avoid using a waiting database for a single-message-only transport.
1589 Performance patch from Paul Fisher. Bugzilla 1262.
1591 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1594 JH/09 Add $headers_added variable, with content from use of ACL modifier
1595 add_header (but not yet added to the message). Bugzilla 199.
1597 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1598 Pulled from Bugzilla 817 by Wolfgang Breyha.
1600 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1602 (nb: this is the same fix as in Exim 4.80.1)
1604 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1607 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1610 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1612 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1613 advertise SMTP AUTH mechanism to us, instead of a generic
1614 protocol violation error. Also, make Exim more robust to bad
1615 data from the Dovecot auth socket.
1617 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1619 When a queue runner is handling a message, Exim first routes the
1620 recipient addresses, during which it prunes them based on the retry
1621 hints database. After that it attempts to deliver the message to
1622 any remaining recipients. It then updates the hints database using
1625 So if a recipient address works intermittently, it can get repeatedly
1626 deferred at routing time. The retry hints record remains fresh so the
1627 address never reaches the final cutoff time.
1629 This is a fairly common occurrence when a user is bumping up against
1630 their storage quota. Exim had some logic in its local delivery code
1631 to deal with this. However it did not apply to per-recipient defers
1632 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1634 This change adds a proper retry rule check during routing so that the
1635 final cutoff time is checked against the message's age. We only do
1636 this check if there is an address retry record and there is not a
1637 domain retry record; this implies that previous attempts to handle
1638 the address had the retry_use_local_parts option turned on. We use
1639 this as an approximation for the destination being like a local
1640 delivery, as in LMTP.
1642 I suspect this new check makes the old local delivery cutoff check
1643 redundant, but I have not verified this so I left the code in place.
1645 TF/02 Correct gecos expansion when From: is a prefix of the username.
1647 Test 0254 submits a message to Exim with the header
1651 When I ran the test suite under the user fanf2, Exim expanded
1652 the header to contain my full name, whereas it should have added
1653 a Resent-Sender: header. It erroneously treats any prefix of the
1654 username as equal to the username.
1656 This change corrects that bug.
1658 GF/01 DCC debug and logging tidyup
1659 Error conditions log to paniclog rather than rejectlog.
1660 Debug lines prefixed by "DCC: " to remove any ambiguity.
1662 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1664 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1665 Bug spotted by Jeremy Harris; was flawed since initial commit.
1666 Would have resulted in OCSP responses post-SNI triggering an Exim
1667 NULL dereference and crash.
1669 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1671 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1672 Bug detection, analysis and fix by Samuel Thibault.
1673 Bugzilla 1331, Debian bug #698092.
1675 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1677 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1678 Server implementation by Todd Lyons, client by JH.
1679 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1680 config variable "prdr_enable" controls whether the server
1681 advertises the facility. If the client requests PRDR a new
1682 acl_data_smtp_prdr ACL is called once for each recipient, after
1683 the body content is received and before the acl_smtp_data ACL.
1684 The client is controlled by both of: a hosts_try_prdr option
1685 on the smtp transport, and the server advertisement.
1686 Default client logging of deliveries and rejections involving
1687 PRDR are flagged with the string "PRDR".
1689 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1690 fclose(). Diagnosis by Todd Lyons.
1692 PP/17 Update configure.default to handle IPv6 localhost better.
1693 Patch by Alain Williams (plus minor tweaks).
1696 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1697 This is now consistent with GnuTLS, and is now documented: the
1698 previous undocumented portable approach to treating the option as
1699 unset was to force an expansion failure. That still works, and
1700 an empty string is now equivalent.
1702 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1703 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1704 not performing validation itself.
1706 PP/20 Added force_command boolean option to pipe transport.
1707 Patch from Nick Koston, of cPanel Inc.
1709 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1712 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1714 PP/21 Fix eximon continuous updating with timestamped log-files.
1715 Broken in a format-string cleanup in 4.80, missed when I repaired the
1716 other false fix of the same issue.
1717 Report and fix from Heiko Schlichting.
1720 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1721 Report from Prashanth Katuri.
1723 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1724 It's SecureTransport, so affects any MacOS clients which use the
1725 system-integrated TLS libraries, including email clients.
1727 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1728 using a MIME ACL for non-SMTP local injection.
1729 Report and assistance in diagnosis by Warren Baker.
1731 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1733 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1735 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1736 last id that failed so it may be referenced in subsequent ACL's.
1738 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1741 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1742 ldap library initialization, allowing self-signed CA's to be
1743 used. Also properly sets require_cert option later in code by
1744 using NULL (global ldap config) instead of ldap handle (per
1745 session). Bug diagnosis and testing by alxgomz.
1747 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1748 the src/util/ subdirectory.
1750 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1751 renamed to Transport Post Delivery Action by Jeremy Harris, as
1754 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1755 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1756 redis_servers = needs to be configured which will be used by the redis
1757 lookup. Patch from Warren Baker, of The Packet Hub.
1759 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1761 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1762 hostname or reverse DNS when processing a host list. Used suggestions
1763 from multiple comments on this bug.
1765 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1767 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1768 Missed a few lines, added it to make the runtest require no keyboard
1771 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1772 contains upper case chars. Make router use caseful_local_part.
1774 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1775 support when GnuTLS has been built with p11-kit.
1781 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1783 This, or similar/improved, will also be change PP/11 of 4.82.
1789 PP/01 Handle short writes when writing local log-files.
1790 In practice, only affects FreeBSD (8 onwards).
1791 Bugzilla 1053, with thanks to Dmitry Isaikin.
1793 NM/01 Bugzilla 949 - Documentation tweak
1795 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1798 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1800 PP/02 Implemented gsasl authenticator.
1802 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1804 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1805 `pkg-config foo` for cflags/libs.
1807 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1808 with rest of GSASL and with heimdal_gssapi.
1810 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1811 `pkg-config foo` for cflags/libs for the TLS implementation.
1813 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1814 properties get this fed in as external SSF. A number of robustness
1815 and debugging improvements to the cyrus_sasl authenticator.
1817 PP/08 cyrus_sasl server now expands the server_realm option.
1819 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1820 Patch by Jeremy Harris.
1822 PP/10 Added dbmjz lookup type.
1824 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1826 PP/12 MAIL args handles TAB as well as SP, for better interop with
1827 non-compliant senders.
1828 Analysis and variant patch by Todd Lyons.
1830 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1831 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1832 Patch from Dirk Mueller <dmueller@suse.com>
1834 PP/13 tls_peerdn now print-escaped for spool files.
1835 Observed some $tls_peerdn in wild which contained \n, which resulted
1836 in spool file corruption.
1838 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1839 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1840 or write after TLS renegotiation, which otherwise led to messages
1843 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1844 as a tracking header (ie: a signed header comes before the signature).
1845 Patch from Wolfgang Breyha.
1847 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1848 comma-sep list; embedded commas doubled.
1850 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1852 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1854 Report and patch from Dmitry Banschikov.
1856 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1857 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1858 failures. We appear to now support TLS1.1+ with Exim.
1860 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1861 lets Exim select keys and certificates based upon TLS SNI from client.
1862 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1863 before an outbound SMTP session. New log_selector, +tls_sni.
1865 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1866 NULL dereference. Report and patch from Alun Jones.
1868 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1869 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1870 Not seeing resolver debug output on NetBSD, but suspect this is a
1871 resolver implementation change.
1873 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1874 Left warnings. Added "eximon gdb" invocation mode.
1876 PP/21 Defaulting "accept_8bitmime" to true, not false.
1878 PP/22 Added -bw for inetd wait mode support.
1880 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1881 locate the relevant includes and libraries. Made this the default.
1883 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1884 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1886 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1887 This may cause build issues on older platforms.
1889 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1890 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1891 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1892 Added SNI support via GnuTLS too.
1893 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1895 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1897 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1898 Applied second patch from Janne, implementing suggestion to default
1899 multiple-strings-in-record handling to match SPF spec.
1901 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1903 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1904 read-only, out of scope).
1905 Patch from Wolfgang Breyha, report from Stuart Northfield.
1907 PP/29 Fix three issues highlighted by clang analyser static analysis.
1908 Only crash-plausible issue would require the Cambridge-specific
1909 iplookup router and a misconfiguration.
1910 Report from Marcin Mirosław.
1912 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1914 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1915 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1916 As part of this, removing so much warning spew let me fix some minor
1917 real issues in debug logging.
1919 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1920 assignment on my part. Fixed.
1922 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1923 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1924 Janne Snabb (who went above and beyond: thank you).
1926 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1927 string otherwise requires a connection and a bunch more work and it's
1928 relatively easy to get wrong. Should also expose TLS library linkage
1931 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1932 64-bit ${eval} (JH/03).
1934 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1935 GNU libc to support some of the 64-bit stuff, should not lead to
1936 conflicts. Defined before os.h is pulled in, so if a given platform
1937 needs to override this, it can.
1939 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1940 protection layer was required, which is not implemented.
1941 Bugzilla 1254, patch from Wolfgang Breyha.
1943 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1944 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1945 tls_dhparam take prime identifiers. Also unbreak combination of
1946 OpenSSL+DH_params+TLSSNI.
1948 PP/39 Disable SSLv2 by default in OpenSSL support.
1954 PP/01 Solaris build fix for Oracle's LDAP libraries.
1955 Bugzilla 1109, patch from Stephen Usher.
1957 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1959 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1962 TF/02 Fix a couple more cases where we did not log the error message
1963 when unlink() failed. See also change 4.74-TF/03.
1965 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1966 lock up or crash if it happened to be inside a call to libc when it
1967 got a SIGUSR1 from exiwhat.
1969 The SIGUSR1 handler appends the current process status to the process
1970 log which is later printed by exiwhat. It used to use the general
1971 purpose logging code to do this, but several functions it calls are
1972 not safe for signals.
1974 The new output code in the SIGUSR1 handler is specific to the process
1975 log, and simple enough that it's easy to inspect for signal safety.
1976 Removing some special cases also simplifies the general logging code.
1977 Removing the spurious timestamps from the process log simplifies
1980 TF/04 Improved ratelimit ACL condition.
1982 The /noupdate option has been deprecated in favour of /readonly which
1983 has clearer semantics. The /leaky, /strict, and /readonly update modes
1984 are mutually exclusive. The update mode is no longer included in the
1985 database key; it just determines when the database is updated. (This
1986 means that when you upgrade Exim will forget old rate measurements.)
1988 Exim now checks that the per_* options are used with an update mode that
1989 makes sense for the current ACL. For example, when Exim is processing a
1990 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1991 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1992 must specify per_mail/readonly. If you omit the update mode it defaults to
1993 /leaky where that makes sense (as before) or /readonly where required.
1995 The /noupdate option is now undocumented but still supported for
1996 backwards compatibility. It is equivalent to /readonly except that in
1997 ACLs where /readonly is required you may specify /leaky/noupdate or
1998 /strict/noupdate which are treated the same as /readonly.
2000 A useful new feature is the /count= option. This is a generalization
2001 of the per_byte option, so that you can measure the throughput of other
2002 aggregate values. For example, the per_byte option is now equivalent
2003 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2005 The per_rcpt option has been generalized using the /count= mechanism
2006 (though it's more complicated than the per_byte equivalence). When it is
2007 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2008 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2009 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2010 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2011 Note that using per_rcpt with a non-readonly update mode in more than
2012 one ACL will cause the recipients to be double-counted. (The per_mail
2013 and per_byte options don't have this problem.)
2015 The handling of very low rates has changed slightly. If the computed rate
2016 is less than the event's count (usually one) then this event is the first
2017 after a long gap. In this case the rate is set to the same as this event's
2018 count, so that the first message of a spam run is counted properly.
2020 The major new feature is a mechanism for counting the rate of unique
2021 events. The new per_addr option counts the number of different
2022 recipients that someone has sent messages to in the last time period. It
2023 behaves like per_rcpt if all the recipient addresses are different, but
2024 duplicate recipient addresses do not increase the measured rate. Like
2025 the /count= option this is a general mechanism, so the per_addr option
2026 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2027 example, measure the rate that a client uses different sender addresses
2028 with the options per_mail/unique=$sender_address. There are further
2029 details in the main documentation.
2031 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2033 TF/06 Removed a few PCRE remnants.
2035 TF/07 Automatically extract Exim's version number from tags in the git
2036 repository when doing development or release builds.
2038 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2039 Bugzilla 879. Patch from Paul Fisher.
2041 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2042 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2045 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2047 PP/05 Variable $av_failed, true if the AV scanner deferred.
2048 Bugzilla 1078. Patch from John Horne.
2050 PP/06 Stop make process more reliably on build failure.
2051 Bugzilla 1087. Patch from Heiko Schlittermann.
2053 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2054 Bugzilla 1089. Patch from Heiko Schlittermann.
2056 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2057 Bugzilla 1131. Patch from Holger Weiß.
2059 PP/09 Handle IPv6 addresses with SPF.
2060 Bugzilla 860. Patch from Wolfgang Breyha.
2062 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2064 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2067 PP/11 match_* no longer expand right-hand-side by default.
2068 New compile-time build option, EXPAND_LISTMATCH_RHS.
2069 New expansion conditions, "inlist", "inlisti".
2071 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2073 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2075 PP/14 fix log_write() format string regression from TF/03.
2076 Bugzilla 1152. Patch from Dmitry Isaikin.
2082 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2084 PP/02 Harmonised TLS library version reporting; only show if debugging.
2085 Layout now matches that introduced for other libraries in 4.74 PP/03.
2087 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2089 PP/04 New "dns_use_edns0" global option.
2091 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2094 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2095 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2097 TK/01 Updated PolarSSL code to 0.14.2.
2098 Bugzilla 1097. Patch from Andreas Metzler.
2100 PP/07 Catch divide-by-zero in ${eval:...}.
2101 Fixes bugzilla 1102.
2103 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2106 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2107 format-string attack -- SECURITY: remote arbitrary code execution.
2109 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2110 time unintentionally subject to list matching rules, letting the header
2111 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2112 arbitrary string expansion). This allowed for information disclosure.
2114 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2115 INT_MIN/-1 -- value coerced to INT_MAX.
2121 NM/01 Workaround for PCRE version dependency in version reporting
2124 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2125 This fixes portability to compilers other than gcc, notably
2126 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2128 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2129 makefiles for portability to HP-UX and POSIX correctness.
2131 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2132 Also via indented variable definition in the Makefile.
2133 (Debugging by Oliver Heesakkers).
2135 PP/02 Restore caching of spamd results with expanded spamd_address.
2136 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2138 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2139 Improves build reliability. Fix from: Frank Elsner
2141 NM/02 Fix wide character breakage in the rfc2047 coding
2142 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2144 NM/03 Allow underscore in dnslist lookups
2145 Fixes bug 1026. Patch from Graeme Fowler
2147 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2148 Code patches from Adam Ciarcinski of NetBSD.
2150 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2153 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2154 is logged, to avoid truncation. Patch from John Horne.
2156 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2157 Patch from Jakob Hirsch.
2159 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2160 SQL string expansion failure details.
2161 Patch from Andrey Oktyabrski.
2163 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2164 Patch from Simon Arlott.
2166 PP/09 New lookups functionality failed to compile on old gcc which rejects
2167 extern declarations in function scope.
2168 Patch from Oliver Fleischmann
2170 PP/10 Use sig_atomic_t for flags set from signal handlers.
2171 Check getgroups() return and improve debugging.
2172 Fixed developed for diagnosis in bug 927 (which turned out to be
2175 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2176 Patch from Mark Zealey.
2178 PP/12 Bugzilla 1056: Improved spamd server selection.
2179 Patch from Mark Zealey.
2181 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2182 Based on patch from Heiko Schlittermann.
2184 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2185 Patch from Uwe Doering, sign-off by Michael Haardt.
2187 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2188 variable declaration deep within a block. Bug and patch from
2191 PP/15 lookups-Makefile IRIX compatibility coercion.
2193 PP/16 Make DISABLE_DKIM build knob functional.
2195 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2196 Patch by Simon Arlott
2198 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2199 variable argument macros. Our copy now differs from upstream.
2205 TF/01 Failure to get a lock on a hints database can have serious
2206 consequences so log it to the panic log.
2208 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2209 controlled using the smtp_confirmation log selector.
2211 TF/03 Include the error message when we fail to unlink a spool file.
2213 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2214 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2215 for maintaining out-of-tree patches for some time.
2217 PP/01 Bugzilla 139: Documentation and portability issues.
2218 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2219 Handle per-OS dynamic-module compilation flags.
2221 PP/02 Let /dev/null have normal permissions.
2222 The 4.73 fixes were a little too stringent and complained about the
2223 permissions on /dev/null. Exempt it from some checks.
2224 Reported by Andreas M. Kirchwitz.
2226 PP/03 Report version information for many libraries, including
2227 Exim version information for dynamically loaded libraries. Created
2228 version.h, now support a version extension string for distributors
2229 who patch heavily. Dynamic module ABI change.
2231 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2232 privilege escalation vulnerability whereby the Exim run-time user
2233 can cause root to append content of the attacker's choosing to
2236 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2239 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2240 If dropping privileges for untrusted macros, we disabled normal logging
2241 on the basis that it would fail; for the Exim run-time user, this is not
2242 the case, and it resulted in successful deliveries going unlogged.
2243 Fixed. Reported by Andreas Metzler.
2249 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2250 only prepend for the Resent-* case. Fixes regression introduced in
2251 Exim 4.70 by NM/22 for Bugzilla 607.
2253 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2254 increasing numbers of administrators be bitten by this.
2256 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2258 PP/03 Bugzilla 994: added openssl_options main configuration option.
2260 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2262 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2264 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2266 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2267 without a peer certificate, leading to a segfault because of an
2268 assumption that peers always have certificates. Be a little more
2269 paranoid. Problem reported by Martin Tscholak.
2271 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2272 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2273 NB: ClamAV planning to remove STREAM in "middle of 2010".
2274 CL also introduces -bmalware, various -d+acl logging additions and
2275 more caution in buffer sizes.
2277 PP/09 Implemented reverse_ip expansion operator.
2279 PP/10 Bugzilla 937: provide a "debug" ACL control.
2281 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2283 PP/12 Bugzilla 973: Implement --version.
2285 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2287 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2289 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2291 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2292 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2293 ignore trailing whitespace.
2295 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2297 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2300 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2301 Notification from Dr Andrew Aitchison.
2303 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2304 ExtendedDetectionInfo response format.
2305 Notification from John Horne.
2307 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2310 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2311 XSL and documented dependency on system catalogs, with examples of how
2314 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2317 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2318 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2319 configuration file which is writeable by the Exim user or group.
2321 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2322 of configuration files to cover files specified with the -C option if
2323 they are going to be used with root privileges, not just the default
2326 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2327 option (effectively making it always true).
2329 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2330 files to be used while preserving root privileges.
2332 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2333 that rogue child processes cannot use them.
2335 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2336 run-time user, instead of root.
2338 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2339 Exim run-time user without dropping privileges.
2341 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2342 result string, instead of calling string_vformat() twice with the same
2345 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2346 for other users. Others should always drop root privileges if they use
2347 -C on the command line, even for a whitelisted configure file.
2349 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2351 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2357 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2358 $data_path, and $header_path variables; fixed documentation bugs and
2361 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2362 exipick to access non-standard spools, including the "frozen" queue
2365 NM/01 Bugzilla 965: Support mysql stored procedures.
2366 Patch from Alain Williams
2368 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2370 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2371 Patch from Andreas Metzler
2373 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2374 Patch from Kirill Miazine
2376 NM/05 Bugzilla 671: Added umask to procmail example.
2378 JJ/03 installed exipick 20100323.0, fixing doc bug
2380 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2381 directory. Notification and patch from Dan Rosenberg.
2383 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2385 TK/02 Improve log output when DKIM signing operation fails.
2387 MH/01 Treat the transport option dkim_domain as a colon separated
2388 list, not as a single string, and sign the message with each element,
2389 omitting multiple occurences of the same signer.
2391 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2392 Bugzilla 985, 986. Patch by Simon Arlott
2394 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2395 Patch by Simon Arlott
2397 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2398 MBX locking. Notification from Dan Rosenberg.
2404 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2406 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2408 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2410 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2412 NM/04 Bugzilla 915: Flag broken perl installation during build.
2418 TK/01 Added patch by Johannes Berg that expands the main option
2419 "spamd_address" if it starts with a dollar sign.
2421 TK/02 Write list of recipients to X-Envelope-Sender header when building
2422 the mbox-format spool file for content scanning (suggested by Jakob
2425 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2426 (http://www.dcc-servers.net/) support via dccifd. Activated by
2427 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2429 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2430 by Mark Daniel Reidel <mr@df.eu>.
2432 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2433 When building exim an external PCRE library is now needed -
2434 PCRE is a system library on the majority of modern systems.
2435 See entry on PCRE_LIBS in EDITME file.
2437 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2438 conversation. Added nologin parameter to request.
2439 Patch contributed by Kirill Miazine.
2441 TF/01 Do not log submission mode rewrites if they do not change the address.
2443 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2445 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2446 log files in place. Contributed by Roberto Lima.
2448 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2450 TF/03 Bugzilla 615: When checking the local_parts router precondition
2451 after a local_part_suffix or local_part_prefix option, Exim now
2452 does not use the address's named list lookup cache, since this
2453 contains cached lookups for the whole local part.
2455 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2456 Robert Millan. Documentation is in experimental-spec.txt.
2458 TF/04 Bugzilla 668: Fix parallel build (make -j).
2460 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2462 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2463 Patch provided by Jan Srzednicki.
2465 TF/05 Leading white space used to be stripped from $spam_report which
2466 wrecked the formatting. Now it is preserved.
2468 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2469 that they are available at delivery time.
2471 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2473 TF/08 TLS error reporting now respects the incoming_interface and
2474 incoming_port log selectors.
2476 TF/09 Produce a more useful error message if an SMTP transport's hosts
2477 setting expands to an empty string.
2479 NM/06 Bugzilla 744: EXPN did not work under TLS.
2480 Patch provided by Phil Pennock.
2482 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2483 Patch provided by Richard Godbee.
2485 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2486 acl_smtp_notquit, added index entry.
2488 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2489 Patch provided by Eugene Bujak.
2491 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2492 accept(). Patch provided by Maxim Dounin.
2494 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2495 Patch provided by Phil Pennock.
2497 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2499 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2500 Patch provided by Brad "anomie" Jorsch.
2502 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2503 Patch provided by Dean Brooks.
2505 TK/05 Add native DKIM support (does not depend on external libraries).
2507 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2508 Patch provided by Graeme Fowler.
2510 NM/16 Bugzilla 851: Documentation example syntax fix.
2512 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2514 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2517 NM/19 Bugzilla 745: TLS version reporting.
2518 Patch provided by Phil Pennock.
2520 NM/20 Bugzilla 167: bool: condition support.
2521 Patch provided by Phil Pennock.
2523 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2524 clients. Patch provided by Phil Pennock.
2526 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2527 Patch provided by Brad "anomie" Jorsch.
2529 NM/23 Bugzilla 687: Fix misparses in eximstats.
2530 Patch provided by Heiko Schlittermann.
2532 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2533 Patch provided by Heiko Schlittermann.
2535 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2536 plus update to original patch.
2538 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2540 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2541 Patch provided by David Brownlee.
2543 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2545 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2547 NM/30 Bugzilla 888: TLS documentation bugfixes.
2549 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2551 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2552 Unlike the original bugzilla I have changed all shell scripts in src tree.
2554 NM/33 Bugzilla 898: Transport filter timeout fix.
2555 Patch by Todd Rinaldo.
2557 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2558 Patch by Serge Demonchaux.
2560 NM/35 Bugzilla 39: Base64 decode bug fixes.
2561 Patch by Jakob Hirsch.
2563 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2565 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2567 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2569 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2575 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2576 ALT-N's libdkim that I have put here:
2577 http://duncanthrax.net/exim-experimental/
2579 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2580 were called 'true' and it seems that C99 defines that as a reserved
2581 keyword to be used with 'bool' variable types. That means you could
2582 not include C99-style headers which use bools without triggering
2583 build errors in sieve.c.
2585 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2586 as mailq or other aliases. Changed the --help handling significantly
2587 to do whats expected. exim_usage() emits usage/help information.
2589 SC/01 Added the -bylocaldomain option to eximstats.
2591 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2593 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2595 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2601 PH/01 Another patch from the Sieve maintainer.
2603 PH/02 When an IPv6 address is converted to a string for single-key lookup
2604 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2605 dots are used instead of colons so that keys in lsearch files need not
2606 contain colons. This was done some time before quoting was made available
2607 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2608 (notated using the quote facility) so as to distinguish them from IPv4
2609 keys. This meant that lookups for IP addresses in host lists did not work
2610 for iplsearch lookups.
2612 This has been fixed by arranging for IPv6 addresses to be expressed with
2613 colons if the lookup type is iplsearch. This is not incompatible, because
2614 previously such lookups could never work.
2616 The situation is now rather anomalous, since one *can* have colons in
2617 ordinary lsearch keys. However, making the change in all cases is
2618 incompatible and would probably break a number of configurations.
2620 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2623 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2624 conversion specification without a maximum field width, thereby enabling
2625 a rogue spamd server to cause a buffer overflow. While nobody in their
2626 right mind would setup Exim to query an untrusted spamd server, an
2627 attacker that gains access to a server running spamd could potentially
2628 exploit this vulnerability to run arbitrary code as the Exim user.
2630 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2631 $primary_hostname instead of what libspf2 thinks the hosts name is.
2633 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2634 a directory entry by the name of the lookup key. Previously, if a
2635 symlink pointed to a non-existing file or a file in a directory that
2636 Exim lacked permissions to read, a lookup for a key matching that
2637 symlink would fail. Now it is enough that a matching directory entry
2638 exists, symlink or not. (Bugzilla 503.)
2640 PH/03 The body_linecount and body_zerocount variables are now exported in the
2643 PH/04 Added the $dnslist_matched variable.
2645 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2646 This means they are set thereafter only if the connection becomes
2649 PH/06 Added the client_condition to authenticators so that some can be skipped
2650 by clients under certain conditions.
2652 PH/07 The error message for a badly-placed control=no_multiline_responses left
2653 "_responses" off the end of the name.
2655 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2657 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2658 (without spaces) instead of just copying the configuration text.
2660 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2662 PH/11 Added $max_received_linelength.
2664 PH/12 Added +ignore_defer and +include_defer to host lists.
2666 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2667 way in which PCRE > 7.0 is built.
2669 PH/14 Implemented queue_only_load_latch.
2671 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2672 MAIL command. The effect was to mangle the value on 64-bit systems.
2674 PH/16 Another patch from the Sieve maintainer.
2676 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2678 PH/18 If a system quota error occurred while trying to create the file for
2679 a maildir delivery, the message "Mailbox is full" was not appended to the
2680 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2681 only to a quota excession during the actual writing of the file.
2683 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2684 characters?) which causes problems in log lines. The DN values are now
2685 passed through string_printing() before being added to log lines.
2687 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2688 and InterBase are left for another time.)
2690 PH/21 Added message_body_newlines option.
2692 PH/22 Guard against possible overflow in moan_check_errorcopy().
2694 PH/23 POSIX allows open() to be a macro; guard against that.
2696 PH/24 If the recipient of an error message contained an @ in the local part
2697 (suitably quoted, of course), incorrect values were put in $domain and
2698 $local_part during the evaluation of errors_copy.
2704 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2705 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2708 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2709 issue a MAIL command.
2711 PH/02 In an ACL statement such as
2713 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2715 if a client was not listed at all, or was listed with a value other than
2716 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2717 the condition was not true (as it should be), so access was not denied.
2718 The bug was that the ! inversion was incorrectly passed on to the second
2719 item. This has been fixed.
2721 PH/03 Added additional dnslists conditions == and =& which are different from
2722 = and & when the dns lookup returns more than one IP address.
2724 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2725 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2727 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2728 FSYNC, which compiles an option called disable_fsync that allows for
2729 bypassing fsync(). The documentation is heavily laced with warnings.
2731 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2733 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2734 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2735 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2736 including adding "make clean"; (3) Added -fPIC when compiling the test
2737 dynamically loaded module, to get rid of a warning.
2739 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2740 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2741 = 0s. The bug is otherwise harmless.
2743 PH/07 There was a bug in the dovecot authenticator such that the value of
2744 $auth1 could be overwritten, and so not correctly preserved, after a
2745 successful authentication. This usually meant that the value preserved by
2746 the server_setid option was incorrect.
2748 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2750 PH/09 Installed PCRE release 7.0.
2752 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2753 run for batched SMTP input. It is now run at the start of every message
2754 in the batch. While fixing this I discovered that the process information
2755 (output by running exiwhat) was not always getting set for -bs and -bS
2756 input. This is fixed, and it now also says "batched" for BSMTP.
2758 PH/11 Added control=no_pipelining.
2760 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2761 patch, slightly modified), and move the expansion of helo_data till after
2762 the connection is made in the smtp transport (so it can use these
2765 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2767 PH/14 Added log_selector = +pid.
2769 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2771 PH/16 Add ${if forany and ${if forall.
2773 PH/17 Added dsn_from option to vary the From: line in DSNs.
2775 PH/18 Flush SMTP output before performing a callout, unless control =
2776 no_callout_flush is set.
2778 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2779 was true (the default) a successful delivery failed to delete the retry
2780 item, thus causing premature timeout of the address. The bug is now
2783 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2785 PH/21 Long custom messages for fakedefer and fakereject are now split up
2786 into multiline responses in the same way that messages for "deny" and
2787 other ACL rejections are.
2789 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2790 with slight modification.
2792 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2793 draft, changing the syntax and factoring some duplicate code".
2795 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2796 for deliveries of the second and subsequent messages over the same SMTP
2799 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2800 ${reduce, with only minor "tidies".
2802 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2804 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2805 expansion side effects.
2807 PH/27 When a message times out after an over-quota error from an Exim-imposed
2808 quota, the bounce message says "mailbox is full". This message was not
2809 being given when it was a system quota that was exceeded. It now should
2812 MH/03 Made $recipients available in local_scan(). local_scan() already has
2813 better access to the recipient list through recipients_list[], but
2814 $recipients can be useful in postmaster-provided expansion strings.
2816 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2817 in the case of a MAIL command with additional options following the
2818 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2819 were accidentally chopped off.
2821 PH/29 SMTP synchronization checks are implemented when a command is read -
2822 there is a check that no more input is waiting when there shouldn't be
2823 any. However, for some commands, a delay in an ACL can mean that it is
2824 some time before the response is written. In this time, more input might
2825 arrive, invalidly. So now there are extra checks after an ACL has run for
2826 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2827 pipelining has not been advertised.
2829 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2831 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2832 returned in the SMTP response when smtp_return_error_details was false.
2833 This has been fixed.
2835 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2836 instead of the C I/O that was originally supplied, because problems were
2837 reported on Solaris.
2839 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2840 Exim which did not show up earlier: it was assuming that a call to
2841 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2842 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2843 was a macro that became an assignment, so it seemed to work. This has
2844 changed to a proper function call with a void return, hence the compile
2845 error. Exim's code has been fixed.
2847 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2850 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2852 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2854 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2855 because it turns out to be too big or there is a timeout) there is no
2856 "Completed" line in the log. When some messages of this type were
2857 selected by exigrep, they were listed as "not completed". Others were
2858 picked up by some special patterns. I have improved the selection
2859 criteria to be more general.
2861 PH/38 The host_find_failed option in the manualroute router can now be set
2862 to "ignore", to completely ignore a host whose IP address cannot be
2863 found. If all hosts are ignored, the behaviour is controlled by the new
2864 host_all_ignored option.
2866 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2867 homing or because of multiple MX records with /mx) generated more than
2868 one IP address, and the following item turned out to be the local host,
2869 all the secondary addresses of the first item were incorrectly removed
2870 from the list, along with the local host and any following hosts (which
2871 is what is supposed to happen).
2873 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2874 whoever called Exim into the -H file. In the case of the daemon it was
2875 behaving confusingly. When first started, it used values for whoever
2876 started the daemon, but after a SIGHUP it used the Exim user (because it
2877 calls itself on a restart). I have changed the code so that it now always
2880 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2881 message are rejected with the same error (e.g. no authentication or bad
2882 sender address), and a DATA command is nevertheless sent (as can happen
2883 with PIPELINING or a stupid MUA), the error message that was given to the
2884 RCPT commands is included in the rejection of the DATA command. This is
2885 intended to be helpful for MUAs that show only the final error to their
2888 PH/42 Another patch from the Sieve maintainer.
2890 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2891 Eximstats - Fixed some broken HTML links and added missing column headers
2893 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2894 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2896 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2898 PH/43 Yet another patch from the Sieve maintainer.
2900 PH/44 I found a way to check for a TCP/IP connection going away before sending
2901 the response to the final '.' that terminates a message, but only in the
2902 case where the client has not sent further data following the '.'
2903 (unfortunately, this is allowed). However, in many cases there won't be
2904 any further data because there won't be any more messages to send. A call
2905 to select() can be used: if it shows that the input is "ready", there is
2906 either input waiting, or the socket has been closed. An attempt to read
2907 the next input character can distinguish the two cases. Previously, Exim
2908 would have sent an OK response which the client would never have see.
2909 This could lead to message repetition. This fix should cure that, at
2910 least in a lot of common cases.
2912 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2913 advertised in response to EHLO.
2919 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2920 fixed by 4.65/MH/01 (is this a record?) are fixed:
2922 (i) An empty string was always treated as zero by the numeric comparison
2923 operators. This behaviour has been restored.
2925 (ii) It is documented that the numeric comparison operators always treat
2926 their arguments as decimal numbers. This was broken in that numbers
2927 starting with 0 were being interpreted as octal.
2929 While fixing these problems I realized that there was another issue that
2930 hadn't been noticed. Values of message_size_limit (both the global option
2931 and the transport option) were treated as octal if they started with 0.
2932 The documentation was vague. These values are now always treated as
2933 decimal, and I will make that clear in the documentation.
2939 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2940 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2943 MH/01 Don't check that the operands of numeric comparison operators are
2944 integers when their expansion is in "skipping" mode (fixes bug
2945 introduced by 4.64-PH/07).
2947 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2948 child addresses, Exim now panics and dies. Previously, because the count
2949 is held in a short int, deliveries were likely to be lost. As such a
2950 large number of recipients for a single message is ridiculous
2951 (performance will be very, very poor), I have chosen to impose a limit
2952 rather than extend the field.
2958 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2959 leftover -K file (the existence of which was triggered by #402).
2960 While we were at it, introduced process PID as part of the -K
2961 filename. This should rule out race conditions when creating
2964 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2965 processing considerably. Previous code took too long for large mails,
2966 triggering a timeout which in turn triggers #401.
2968 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2969 in the DK code in transports.c. sendfile() is not really portable,
2970 hence the _LINUX specificness.
2972 TF/01 In the add_headers option to the mail command in an Exim filter,
2973 there was a bug that Exim would claim a syntax error in any
2974 header after the first one which had an odd number of characters
2977 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2978 callout verification, Exim cached a "reject" for the entire domain. This
2979 is correct for most verifications, but it is not correct for a recipient
2980 verification with use_sender or use_postmaster set, because in that case
2981 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2982 case of MAIL FROM:<> rejection from other early rejections (e.g.
2983 rejection of HELO). When verifying a recipient using a non-null MAIL
2984 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2985 Whatever the result of the callout, the value of the domain cache is
2986 left unchanged (for any other kind of callout, getting as far as trying
2987 RCPT means that the domain itself is ok).
2989 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2992 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2993 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2994 session, the daemon ignores ECONNECT errors and logs others; it now
2995 ignores EPIPE as well.
2997 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2998 (quoted-printable decoding).
3000 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3001 later the small subsequent patch to fix an introduced bug.
3003 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3005 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3007 PH/08 An error is now given if message_size_limit is specified negative.
3009 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3010 to be given (somewhat) arbitrary names.
3012 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3015 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3016 miscellaneous code fixes
3018 PH/10 Added the log_reject_target ACL modifier to specify where to log
3021 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3022 hostname. This is wrong, because it relates to the incoming message (and
3023 probably the interface on which it is arriving) and not to the outgoing
3024 callout (which could be using a different interface). This has been
3025 changed to use the value of the helo_data option from the smtp transport
3026 instead - this is what is used when a message is actually being sent. If
3027 there is no remote transport (possible with a router that sets up host
3028 addresses), $smtp_active_hostname is used.
3030 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3031 tweaks were necessary in order to get it to work (see also 21 below):
3032 (a) The code assumed that strncpy() returns a negative number on buffer
3033 overflow, which isn't the case. Replaced with Exim's string_format()
3035 (b) There were several signed/unsigned issues. I just did the minimum
3036 hacking in of casts. There is scope for a larger refactoring.
3037 (c) The code used strcasecmp() which is not a standard C function.
3038 Replaced with Exim's strcmpic() function.
3039 (d) The code set only $1; it now sets $auth1 as well.
3040 (e) A simple test gave the error "authentication client didn't specify
3041 service in request". It would seem that Dovecot has changed its
3042 interface. Fortunately there's a specification; I followed it and
3043 changed what the client sends and it appears to be working now.
3045 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3048 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3049 address A is aliased to B and C, where B exists and C does not. Without
3050 -v the output is "A verified" because verification stops after a
3051 successful redirection if more than one address is generated. However,
3052 with -v the child addresses are also verified. Exim was outputting "A
3053 failed to verify" and then showing the successful verification for C,
3054 with its parentage. It now outputs "B failed to verify", showing B's
3055 parentage before showing the successful verification of C.
3057 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3058 look up a TXT record in a specific list after matching in a combined
3061 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3062 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3063 they consult the DNS. I had assumed they would set it the way they
3064 wanted; and indeed my experiments on Linux seem to show that in some
3065 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3066 To be on the safe side, however, I have now made the interface to
3067 host_find_byname() similar to host_find_bydns(), with an argument
3068 containing the DNS resolver options. The host_find_byname() function now
3069 sets these options at its start, just as host_find_bydns() does. The smtp
3070 transport options dns_qualify_single and dns_search_parents are passed to
3071 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3072 of host_find_byname() use the default settings of RES_DEFNAMES
3073 (qualify_single) but not RES_DNSRCH (search_parents).
3075 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3076 spool_read_header() do less string testing, by means of a preliminary
3077 switch on the second character of optional "-foo" lines. (This is
3078 overdue, caused by the large number of possibilities that now exist.
3079 Originally there were few.) While I was there, I also converted the
3080 str(n)cmp tests so they don't re-test the leading "-" and the first
3081 character, in the hope this might squeeze out yet more improvement.
3083 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3084 flag allowing group syntax was set by the header_syntax check but not
3085 turned off, possible causing trouble later; (2) The flag was not being
3086 set at all for the header_verify test, causing "group"-style headers to
3087 be rejected. I have now set it in this case, and also caused header_
3088 verify to ignore an empty address taken from a group. While doing this, I
3089 came across some other cases where the code for allowing group syntax
3090 while scanning a header line wasn't quite right (mostly, not resetting
3091 the flag correctly in the right place). These bugs could have caused
3092 trouble for malformed header lines. I hope it is now all correct.
3094 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3095 with the "reply" argument non-NULL. The code, however (which originally
3096 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3097 but it didn't always do it. This confused somebody who was copying the
3098 code for some other use. I have removed all the tests.
3100 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3101 feature that was used to support insecure browsers during the U.S. crypto
3102 embargo. It requires special client support, and Exim is probably the
3103 only MTA that supported it -- and would never use it because real RSA is
3104 always available. This code has been removed, because it had the bad
3105 effect of slowing Exim down by computing (never used) parameters for the
3106 RSA_EXPORT functionality.
3108 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3109 authenticator to fail if there's a tab character in the incoming data
3110 (there should never be unless someone is messing about, as it's supposed
3111 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3112 if the connection is using TLS or if the remote IP is the same as the
3113 local IP, and the "valid-client-cert option" if a client certificate has
3116 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3117 authenticators. This can be used for authorization after authentication
3118 succeeds. (In the case of plaintext, it servers for both authentication
3121 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3122 if any retry times were supplied.
3124 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3125 connection, where there is no client IP address to check. In this
3126 situation, the verify now always succeeds.
3128 PH/25 Applied John Jetmore's -Mset patch.
3130 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3132 PH/27 In a string expansion for a processed (not raw) header when multiple
3133 headers of the same name were present, leading whitespace was being
3134 removed from all of them, but trailing whitespace was being removed only
3135 from the last one. Now trailing whitespace is removed from each header
3136 before concatenation. Completely empty headers in a concatenation (as
3137 before) are ignored.
3139 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3140 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3142 PH/29 [Removed. This was a change that I later backed out, and forgot to
3143 correct the ChangeLog entry (that I had efficiently created) before
3144 committing the later change.]
3146 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3147 address errors (4xx response to RCPT) over the same connection as other
3148 messages routed to the same hosts. Such deliveries are always "forced",
3149 so retry times are not inspected. This resulted in far too many retries
3150 for the affected addresses. The effect occurred only when there were more
3151 hosts than the hosts_max_try setting in the smtp transport when it had
3152 the 4xx errors. Those hosts that it had tried were not added to the list
3153 of hosts for which the message was waiting, so if all were tried, there
3154 was no problem. Two fixes have been applied:
3156 (i) If there are any address or message errors in an SMTP delivery, none
3157 of the hosts (tried or untried) are now added to the list of hosts
3158 for which the message is waiting, so the message should not be a
3159 candidate for sending over the same connection that was used for a
3160 successful delivery of some other message. This seems entirely
3161 reasonable: after all the message is NOT "waiting for some host".
3162 This is so "obvious" that I'm not sure why it wasn't done
3163 previously. Hope I haven't missed anything, but it can't do any
3164 harm, as the worst effect is to miss an optimization.
3166 (ii) If, despite (i), such a delivery is accidentally attempted, the
3167 routing retry time is respected, so at least it doesn't keep
3168 hammering the server.
3170 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3171 in ${readsocket because some servers need this prod.
3173 PH/32 Added some extra debug output when updating a wait-xxx database.
3175 PH/33 The hint "could be header name not terminated by colon", which has been
3176 given for certain expansion errors for a long time, was not being given
3177 for the ${if def:h_colon_omitted{... case.
3179 PH/34 The spec says: "With one important exception, whenever a domain list is
3180 being scanned, $domain contains the subject domain." There was at least
3181 one case where this was not true.
3183 PH/35 The error "getsockname() failed: connection reset by peer" was being
3184 written to the panic log as well as the main log, but it isn't really
3185 panic-worthy as it just means the connection died rather early on. I have
3186 removed the panic log writing for the ECONNRESET error when getsockname()
3189 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3190 runs only) independently of the message's sender address. This meant
3191 that, if the 4xx error was in fact related to the sender, a different
3192 message to the same recipient with a different sender could confuse
3193 things. In particular, this can happen when sending to a greylisting
3194 server, but other circumstances could also provoke similar problems.
3195 I have changed the default so that the retry time for these errors is now
3196 based a combination of the sender and recipient addresses. This change
3197 can be overridden by setting address_retry_include_sender=false in the
3200 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3201 remote server are returned as part of bounce messages. This was not
3202 happening for LMTP over a pipe (the lmtp transport), but now it is the
3203 same for both kinds of LMTP.
3205 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3206 in header lines that were in fact CNAMEs. This is no longer the case.
3208 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3209 and queue runs started by the daemon processed all messages. This has
3210 been fixed so that -R and -S can now usefully be given with -q<time>.
3212 PH/40 Import PCRE release 6.7 (fixes some bugs).
3214 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3216 PH/42 Give an error if -q is specified more than once.
3218 PH/43 Renamed the variables $interface_address and $interface_port as
3219 $received_ip_address and $received_port, to make it clear that these
3220 values apply to message reception, and not to the outgoing interface when
3221 a message is delivered. (The old names remain recognized, of course.)
3223 PH/44 There was no timeout on the connect() call when using a Unix domain
3224 socket in the ${readsocket expansion. There now is.
3226 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3227 be meaningful with "accept".
3229 SC/01 Eximstats V1.43
3230 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3232 SC/02 Eximstats V1.44
3233 Use a glob alias rather than an array ref in the generated
3234 parser. This improves both readability and performance.
3236 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3237 Collect SpamAssassin and rejection statistics.
3238 Don't display local sender or destination tables unless
3239 there is data to show.
3240 Added average volumes into the top table text output.
3242 SC/04 Eximstats V1.46
3243 Collect data on the number of addresses (recipients)
3244 as well as the number of messages.
3246 SC/05 Eximstats V1.47
3247 Added 'Message too big' to the list of mail rejection
3248 reasons (thanks to Marco Gaiarin).
3250 SC/06 Eximstats V1.48
3251 Mainlog lines which have GMT offsets and are too short to
3252 have a flag are now skipped.
3254 SC/07 Eximstats V1.49 (Alain Williams)
3255 Added the -emptyok flag.
3257 SC/08 Eximstats V1.50
3258 Fixes for obtaining the IP address from reject messages.
3260 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3261 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3262 whitespace changes from 4.64-PH/27
3264 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3267 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3268 are found, allow negative numbers in numeric criteria)
3270 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3272 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3273 to match changes made in 4.64-PH/43
3275 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3277 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3278 "domain" is actually an IP address, and force a failure. This locks out
3279 those revolvers/nameservers that support "A-for-A" lookups, in
3280 contravention of the specifications.
3282 PH/48 When a host name was looked up from an IP address, and the subsequent
3283 forward lookup of the name timed out, the host name was left in
3284 $sender_host_name, contrary to the specification.
3286 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3287 restricted to single-key lookups, Exim was not diagnosing an error if
3288 * or *@ was used with a query-style lookup.
3290 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3292 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3293 long ago, but noone interested enough thought of it. Let's just say that
3294 the "1.1" means that there are some new functions that weren't there at
3295 some point in the past.
3297 PH/51 Error processing for expansion failure of helo_data from an smtp
3298 transport during callout processing was broken.
3300 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3301 tested/used via the -bh/-bhc/-bs options.
3303 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3304 bug, fixed in subsequent PCRE releases).
3306 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3307 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3309 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3315 SC/01 Use a glob alias rather than an array ref in eximstats generated
3316 parser. This improves both readability and performance.
3318 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3319 Don't display local sender or destination tables in eximstats unless
3320 there is data to show.
3321 Added average volumes into the eximstats top table text output.
3323 SC/03 Collect data on the number of addresses (recipients) as well
3324 as the number of messages in eximstats.
3326 TF/01 Correct an error in the documentation for the redirect router. Exim
3327 does (usually) call initgroups() when daemonizing.
3329 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3330 with consistent privilege compared to when running as a daemon.
3332 TF/03 Note in the spec that $authenticated_id is not set for local
3333 submissions from trusted users.
3335 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3336 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3338 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3339 by adding some example configuration directives to the default
3340 configuration file. A little bit of work is required to uncomment the
3341 directives and define how usernames and passwords are checked, but
3342 there is now a framework to start from.
3344 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3345 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3346 without this. I don't know how relevant this is to other LDAP libraries.
3348 PH/02 Add the verb name to the "unknown ACL verb" error.
3350 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3352 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3354 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3355 directory not expanded when it should be if an expanded home directory
3356 was set for the address (which is overridden by the transport).
3358 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3361 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3362 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3363 because it is too late at that time, and has no effect.
3365 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3366 security issue with \' (bugzilla #107). I could not use the
3367 PQescapeStringConn() function, because it needs a PGconn value as one of
3370 PH/08 When testing addresses using -bt, indicate those final addresses that
3371 are duplicates that would not cause an additional delivery. At least one
3372 person was confused, thinking that -bt output corresponded to deliveries.
3373 (Suppressing duplicates isn't a good idea as you lose the information
3374 about possibly different redirections that led to the duplicates.)
3376 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3377 systems where poll() doesn't work, in particular OS X.
3379 PH/10 Added more information to debugging output for retry time not reached.
3381 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3382 operations in malware.c.
3384 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3387 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3388 syslog_duplication set false, log lines that would normally be written
3389 both the the main log and to the reject log were not written to syslog at
3392 PH/14 In the default configuration, change the use of "message" in ACL warn
3393 statements to "add_header".
3395 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3396 not followed by a command (e.g. "seen endif").
3398 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3399 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3402 PH/17 Added extra conditions to the default value of delay_warning_condition
3406 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3407 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3408 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3411 The Auto-Submitted: and various List- headers are standardised, whereas I
3412 don't think Precedence: ever was.
3414 PH/18 Refactored debugging code in route_finduser() to show more information,
3415 in particular, the error code if getpwnam() issues one.
3417 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3418 This is apparently needed in addition to the PH/07 change above to avoid
3419 any possible encoding problems.
3421 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3422 but not after initializing Perl.
3424 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3425 output them only if debugging. By default they are written stderr,
3426 apparently, which is not desirable.
3428 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3431 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3434 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3436 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3437 authenticated or an ident call has been made. Suppress the default
3438 values for $authenticated_id and $authenticated_sender (but permit -oMai
3439 and -oMas) when testing with -bh.
3441 PH/24 Re-jigged the order of the tests in the default configuration so that the
3442 tests for valid domains and recipients precede the DNS black list and CSA
3443 tests, on the grounds that those ones are more expensive.
3445 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3446 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3447 This bug exists in every version of Exim that I still have, right back to
3450 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3451 However, an attempt to turn on case-sensitivity in a regex key by
3452 including (?-i) didn't work because the subject string was already
3453 lowercased, and the effects were non-intuitive. It turns out that a
3454 one-line patch can be used to allow (?-i) to work as expected.
3460 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3461 other effects) broke the use of negated acl sub-conditions.
3463 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3466 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3467 "Deny" causes Exim to reject the incoming connection with a 554 error.
3468 Unfortunately, if there is a major crisis, such as a disk failure,
3469 tcp-wrappers gives "deny", whereas what one would like would be some
3470 kind of temporary error. A kludge has been added to help with this.
3471 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3472 554 error is used if errno is still zero or contains ENOENT (which occurs
3473 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3476 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3478 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3479 errors. Otherwise a message that provokes a temporary error (when other
3480 messages do not) can cause a whole host to time out.
3482 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3483 addresses were routed directly to files or pipes from a redirect router.
3484 File deliveries just didn't batch; pipe deliveries might have suffered
3487 PH/06 A failure to get a lock for a hints database would erroneously always say
3488 "Failed to get write lock", even when it was really a read lock.
3490 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3491 of 0600. This has been changed to use the value of the lockfile_mode
3492 option (which defaults to 0600).
3494 PH/08 Applied small patch from the Sieve maintainer.
3496 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3497 folder from quota calculations, a direct delivery into this folder messed
3498 up the contents of the maildirsize file. This was because the regex was
3499 used only to exclude .Trash (or whatever) when the size of the mailbox
3500 was calculated. There was no check that a delivery was happening into an
3501 excluded directory. This bug has been fixed by ignoring all quota
3502 processing for deliveries into excluded directories.
3504 PH/10 Added the maildirfolder_create_regex option to appendfile.
3510 PH/01 The code for finding all the local interface addresses on a FreeBSD
3511 system running IPv6 was broken. This may well have applied to all BSD
3512 systems, as well as to others that have similar system calls. The broken
3513 code found IPv4 interfaces correctly, but gave incorrect values for the
3514 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3515 that it would not match correctly against @[] and not recognize the IPv6
3518 PH/02 The ipliteral router was not recognizing addresses of the form user@
3519 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3521 PH/03 Added disable_ipv6.
3523 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3524 decoded form, because it is most often used to construct To: headers
3525 lines in autoreplies, and the decoded form may well be syntactically
3526 invalid. However, $reply_address has leading white space removed, and all
3527 newlines turned into spaces so that the autoreply transport does not
3530 PH/05 If group was specified without a user on a router, and no group or user
3531 was specified on a transport, the group from the router was ignored.
3533 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3534 for visible compile-time settings that can be used to change these
3535 numbers, for those that want even more. Backwards compatibility with old
3536 spool files has been maintained. However, going back to a previous Exim
3537 release will lost any variables that are in spool files.
3539 PH/07 Two small changes when running in the test harness: increase delay when
3540 passing a TCP/IP connection to a new process, in case the original
3541 process has to generate a bounce, and remove special handling of
3542 127.0.0.2 (sic), which is no longer necessary.
3544 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3545 be the same on different OS.
3547 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3550 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3551 whether --show-vars was specified or not
3553 JJ/02 exipick: Added support for new ACL variable spool format introduced
3556 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3557 syntactically invalid From: or Reply-to: line, and a filter used this to
3558 generate an autoreply, and therefore failed to obtain an address for the
3559 autoreply, Exim could try to deliver to a non-existent relative file
3560 name, causing unrelated and misleading errors. What now happens is that
3561 it logs this as a hard delivery error, but does not attempt to create a
3564 PH/11 The exinext utility has a -C option for testing purposes, but although
3565 the given file was scanned by exinext itself; it wasn't being passed on
3566 when Exim was called.
3568 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3569 an end-of-file indication when reading a command response.
3571 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3572 compiled. In many other places in Exim, IPv6 addresses are always
3573 recognized, so I have changed this. It also means that IPv4 domain
3574 literals of the form [IPV4:n.n.n.n] are now always recognized.
3576 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3577 used if the router is not running as root, for example, when verifying at
3578 ACL time, or when using -bh. The debugging output from this situation was
3579 non-existent - all you got was a failure to exec. I have made two
3582 (a) Failures to set uid/gid, the current directory, or a process leader
3583 in a subprocess such as that created by queryprogram now generate
3584 suitable debugging output when -d is set.
3586 (b) The queryprogram router detects when it is not running as root,
3587 outputs suitable debugging information if -d is set, and then runs
3588 the subprocess without attempting to change uid/gid.
3590 PH/15 Minor change to Makefile for building test_host (undocumented testing
3593 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3594 additional section of a DNS packet that returns MX or SRV records.
3595 Instead, it always explicitly searches for A/AAAA records. This avoids
3596 major problems that occur when a DNS server includes only records of one
3597 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3598 fixed another bug: if SRV records were looked up and the corresponding
3599 address records were *not* found in the additional section, the port
3600 values from the SRV records were lost.
3602 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3603 using the correct key (the original address) when searching the retry
3604 rules in order to find which one to use for generating the retry hint.
3606 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3607 adding the default one. Similarly, if it contains a Reply-To: header, the
3608 errors_reply_to option, if set, is not used.
3610 PH/19 When calculating a retry time, Exim used to measure the "time since
3611 failure" by looking at the "first failed" field in the retry record. Now
3612 it does not use this if it is later than than the arrival time of the
3613 message. Instead it uses the arrival time. This makes for better
3614 behaviour in cases where some deliveries succeed, thus re-setting the
3615 "first failed" field. An example is a quota failure for a huge message
3616 when small messages continue to be delivered. Without this change, the
3617 "time since failure" will always be short, possible causing more frequent
3618 delivery attempts for the huge message than are intended.
3619 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3621 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3622 $1, $2, $3) because the numerical variables can be reset during some
3623 expansion items (e.g. "match"), thereby losing the authentication data.
3625 PH/21 Make -bV show the size of off_t variables so that the test suite can
3626 decide whether to run tests for quotas > 2G.
3628 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3629 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3630 filecount value is greater than 2G or if a quota value is greater than 2G
3631 on a system where the size of off_t is not greater than 4, a panic error
3634 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3635 never match. The debug and -bh output now contains an explicit error
3636 message indicating a malformed IPv4 address or mask.
3638 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3639 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3640 PH/23 above applies.
3642 PH/25 Do not write to syslog when running in the test harness. The only
3643 occasion when this arises is a failure to open the main or panic logs
3644 (for which there is an explicit test).
3646 PH/26 Added the /no_tell option to "control=freeze".
3648 PH/27 If a host name lookup failed very early in a connection, for example, if
3649 the IP address matched host_lookup and the reverse lookup yielded a name
3650 that did not have a forward lookup, an error message of the form "no IP
3651 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3652 could be logged. Now it outputs the IP address instead of "NULL".
3654 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3655 allows the sender and the authenticated sender to be set when
3656 submitting a message from within Exim. Since child_open_exim() is
3657 documented for local_scan(), the new function should be too.
3659 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3660 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3661 results in an empty string is now treated as unset.
3663 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3665 PH/31 Added sender_verify_fail logging option.
3667 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3668 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3669 tidied the source and removed it altogether.
3671 PH/33 When a queue run was abandoned because the load average was too high, a
3672 log line was always written; now it is written only if the queue_run log
3673 selector is set. In addition, the log line for abandonment now contains
3674 information about the queue run such as the pid. This is always present
3675 in "start" and "stop" lines but was omitted from the "abandon" line.
3677 PH/34 Omit spaces between a header name and the colon in the error message that
3678 is given when verify = headers_syntax fails (if there are lots of them,
3679 the message gets confusing).
3681 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3682 names, as there are now some PTR records that contain slashes. This check
3683 is only to protect against broken name servers that fall over on strange
3684 characters, so the fact that it applies to all lookups doesn't matter.
3686 PH/36 Now that the new test suite is complete, we can remove some of the
3687 special code in Exim that was needed for the old test suite. For example,
3688 sorting DNS records because real resolvers return them in an arbitrary
3689 order. The new test suite's fake resolver always returns records in the
3692 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3693 bounces) except when queue_only is set, to avoid logging races between
3694 the different processes.
3696 PH/38 Panic-die if .include specifies a non-absolute path.
3698 PH/39 A tweak to the "H" retry rule from its user.
3700 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3701 a label. They prevented compilation on older perls.
3703 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3704 a warning to be raised on newish perls.
3706 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3707 on queue. Changes to match documented behaviour of showing count of
3708 messages matching specified criteria.
3710 PH/40 Changed the default ident timeout from 30s to 5s.
3712 PH/41 Added support for the use of login_cap features, on those BSD systems
3713 that have them, for controlling the resources used by pipe deliveries.
3715 PH/42 The content-scanning code uses fopen() to create files in which to put
3716 message data. Previously it was not paying any attention to the mode of
3717 the files. Exim runs with umask(0) because the rest of the code creates
3718 files with open(), and sets the required mode explicitly. Thus, these
3719 files were ending up world-writeable. This was not a big issue, because,
3720 being within the spool directory, they were not world-accessible. I have
3721 created a function called modefopen, which takes an additional mode
3722 argument. It sets umask(777), creates the file, chmods it to the required
3723 mode, then resets the umask. All the relevant calls to fopen() in the
3724 content scanning code have been changed to use this function.
3726 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3727 to 24 hours. This avoids potential overflow problems when processing G
3728 and H retry rules. I suspect nobody ever tinkers with this value.
3730 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3732 PH/45 When the plaintext authenticator is running as a client, the server's
3733 challenges are checked to ensure they are valid base64 strings. By
3734 default, the authentication attempt is cancelled if an invalid string is
3735 received. Setting client_ignore_invalid_base64 true ignores these errors.
3736 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3737 they are received. Thus, the responses can be made to depend on the
3738 challenges. If an invalid string is ignored, an empty string is placed in
3741 PH/46 Messages that are created by the autoreply transport now contains a
3742 References: header, in accordance with RFCs 2822 and 3834.
3744 PH/47 Added authenticated_sender_force to the smtp transport.
3746 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3748 PH/49 Installed latest patch from the Sieve maintainer.
3750 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3751 was also set, the appendfile transport was unnecessarily scanning a
3752 directory of message files (e.g. for maildir delivery) to find the count
3753 of files (along with the size), even though it did not need this
3754 information. It now does the scan only if it needs to find either the
3755 size of the count of files.
3757 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3759 PH/52 Two bugs concerned with error handling when the smtp transport is
3762 (i) Exim was not creating retry information for temporary errors given
3763 for individual recipients after the DATA command when the smtp transport
3764 was used in LMTP mode. This meant that they could be retried too
3765 frequently, and not timed out correctly.
3767 (ii) Exim was setting the flag that allows error details to be returned
3768 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3769 recipients that were returned after the DATA command.
3771 PH/53 This is related to PH/52, but is more general: for any failing address,
3772 when detailed error information was permitted to be returned to the
3773 sender, but the error was temporary, then after the final timeout, only
3774 "retry timeout exceeded" was returned. Now it returns the full error as
3775 well as "retry timeout exceeded".
3777 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3778 do this, and (what is worse) MTAs that accept it.
3780 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3781 will now be deprecated.
3783 PH/56 New os.c-cygwin from the Cygwin maintainer.
3785 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3786 formats (previously only available in exim formats via -bpr, -bpru,
3787 and -bpra. Now also available in native and exiqgrep formats)
3789 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3790 with very large, slow to parse queues
3792 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3794 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3796 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3797 responses to authentication challenges, though it was showing the
3798 challenges; (ii) I've removed the CR characters from the debug output for
3801 PH/58 Allow for the insertion of a newline as well as a space when a string
3802 is turned into more than one encoded-word during RFC 2047 encoding. The
3803 Sieve code now uses this.
3805 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3806 data_4xx, lost_connection, tls_required.
3808 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3809 message was being sent as an SMTP response.
3811 PH/61 Add -l and -k options to exicyclog.
3813 PH/62 When verifying, if an address was redirected to one new address, so that
3814 verification continued, and the new address failed or deferred after
3815 having set something in $address_data, the value of $address_data was not
3816 passed back to the ACL. This was different to the case when no
3817 redirection occurred. The value is now passed back in both cases.
3819 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3820 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3821 use login_cap.h, so on its own it isn't the distinguishing feature. The
3822 new name refers directly to the setclassresources() function.
3824 PH/65 Added configuration files for NetBSD3.
3826 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3828 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3829 is preferred over IPv4.
3831 PH/68 The bounce_return_message and bounce_return_body options were not being
3832 honoured for bounces generated during the reception of non-SMTP messages.
3833 In particular, this applied to messages rejected by the ACL. This bug has
3834 been fixed. However, if bounce_return_message is true and bounce_return_
3835 body is false, the headers that are returned for a non-SMTP message
3836 include only those that have been read before the error was detected.
3837 (In the case of an ACL rejection, they have all been read.)
3839 PH/69 The HTML version of the specification is now built in a directory called
3840 spec_html instead of spec.html, because the latter looks like a path with
3841 a MIME-type, and this confuses some software.
3843 PH/70 Catch two compiler warnings in sieve.c.
3845 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3846 function verify_get_ident() calls ip_connect() to connect a socket, but
3847 if the "connect()" function timed out, ip_connect() used to close the
3848 socket. However, verify_get_ident() also closes the socket later, and in
3849 between Exim writes to the log, which may get opened at this point. When
3850 the socket was closed in ip_connect(), the log could get the same file
3851 descriptor number as the socket. This naturally causes chaos. The fix is
3852 not to close the socket in ip_connect(); the socket should be closed by
3853 the function that creates it. There was only one place in the code where
3854 this was missing, in the iplookup router, which I don't think anybody now
3855 uses, but I've fixed it anyway.
3857 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3858 well as to direct DNS lookups. Otherwise the handling of names in host
3859 lists is inconsistent and therefore confusing.
3865 PH/01 Two changes to the default runtime configuration:
3867 (1) Move the checks for relay_from_hosts and authenticated clients from
3868 after to before the (commented out) DNS black list checks.
3870 (2) Add control=submission to the relay_from_hosts and authenticated
3871 clients checks, on the grounds that messages accepted by these
3872 statements are most likely to be submissions.
3874 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3876 (1) Generate an error if the third argument for the ${prvs expansion is
3879 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3882 (3) Reset the variables that are obtained from the first argument of
3883 ${prvscheck and used in the second argument before leaving the code,
3884 because their memory is reclaimed, so using them afterwards may do
3887 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3888 one (it's much easier than Tom thought :-).
3890 (5) Because of (4), we can now allow for the use of $prvscheck_result
3891 inside the third argument.
3893 PH/03 For some reason, the default setting of PATH when running a command from
3894 a pipe transport was just "/usr/bin". I have changed it to
3897 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3898 anything to be listed in the output from -bV.
3900 PH/05 When a filter generated an autoreply, the entire To: header line was
3901 quoted in the delivery log line, like this:
3903 => >A.N.Other <ano@some.domain> <original@ddress> ...
3905 This has been changed so that it extracts the operative address. There
3906 may be more than one such address. If so, they are comma-separated, like
3909 => >ano@some.domain,ona@other.domain <original@ddress> ...
3911 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3912 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3913 not being looked up in the rDNS to get a host name, Exim was showing the
3914 IP address twice in Received: lines, even though the IP addresses were
3915 identical. For example:
3917 Received: from [1.2.3.4] (helo=[1.2.3.4])
3919 However, if the real host name was known, it was omitting the HELO data
3920 if it matched the actual IP address. This has been tidied up so that it
3921 doesn't show the same IP address twice.
3923 PH/07 When both +timestamp and +memory debugging was on, the value given by
3924 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3925 called by the debug printing, thereby overwriting the timestamp buffer.
3926 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3928 PH/08 When the original message was included in an autoreply transport, it
3929 always said "this is a copy of the message, including all the headers",
3930 even if body_only or headers_only was set. It now gives an appropriate
3933 PH/09 Applied a patch from the Sieve maintainer which:
3935 o fixes some comments
3936 o adds the (disabled) notify extension core
3937 o adds some debug output for the result of if/elsif tests
3938 o points to the current vacation draft in the documentation
3939 and documents the missing references header update
3943 o fixes a bug in processing the envelope test (when testing
3944 multiple envelope elements, the last element determined the
3947 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3948 Electronic Mail") by including:
3950 Auto-submitted: auto-generated
3952 in the messages that it generates (bounce messages and others, such as
3953 warnings). In the case of bounce messages for non-SMTP messages, there was
3954 also a typo: it was using "Auto_submitted" (underscore instead of
3955 hyphen). Since every message generated by Exim is necessarily in response
3956 to another message, thes have all been changed to:
3958 Auto-Submitted: auto-replied
3960 in accordance with these statements in the RFC:
3962 The auto-replied keyword:
3964 - SHOULD be used on messages sent in direct response to another
3965 message by an automatic process,
3967 - MUST NOT be used on manually-generated messages,
3969 - MAY be used on Delivery Status Notifications (DSNs) and Message
3970 Disposition Notifications (MDNs),
3972 - MUST NOT be used on messages generated by automatic or periodic
3973 processes, except for messages which are automatic responses to
3976 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3977 to the default Received: header definition.
3979 PH/12 Added log selector acl_warn_skipped (default on).
3981 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3982 variables because (a) they are in the wrong storage pool and (b) even if
3983 they were copied, it wouldn't work properly because of the caching.
3985 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3986 checking when decoding. Apparently there are clients that generate
3987 overlong encoded strings. Why am I not surprised?
3989 PH/15 If the first argument of "${if match_address" was not empty, but did not
3990 contain an "@" character, Exim crashed. Now it writes a panic log message
3991 and treats the condition as false.
3993 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3995 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3996 extension "envelope-auth". The code is finished and in agreement with
3997 other implementations, but there is no documentation so far and in fact,
3998 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3999 not changing the active code.
4001 Print executed "if" and "elsif" statements when debugging is used. This
4002 helps a great deal to understand what a filter does.
4004 Document more things not specified clearly in RFC3028. I had all this
4005 sorted out, when out of a sudden new issues came to my mind. Oops."
4007 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4010 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4011 canonical form (as documented). However, after a host name lookup from
4012 the IP address, check_host() was doing a simple string comparison with
4013 addresses acquired from the DNS when checking that the found name did
4014 have the original IP as one of its addresses. Since any found IPv6
4015 addresses are likely to be in abbreviated form, the comparison could
4016 fail. Luckily, there already exists a function for doing the comparison
4017 by converting both addresses to binary, so now that is used instead of
4018 the text comparison.
4020 PH/20 There was another similar case to PH/19, when a complete host name was
4021 given in a host list; looking up its IP address could give an abbreviated
4022 form, whereas the current host's name might or might not be abbreviated.
4023 The same fix has been applied.
4029 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4030 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4033 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4034 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4036 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4038 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4039 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4040 Also, for "Auto-Submitted", treat anything other than "no" as
4041 non-personal, in accordance with RFC 3834. (Previously it treated
4042 anything starting "auto-" as non-personal.)
4044 TF/01 The control=submission/name=... option had a problem with syntax
4045 errors if the name included a slash character. The /name= option
4046 now slurps the rest of the string, so it can include any characters
4047 but it must come last in the list of options (after /sender_retain
4050 PH/05 Some modifications to the interface to the fake nameserver for the new
4058 TK/01 Added the "success_on_redirect" address verification option. See
4059 NewStuff for rationale and an example.
4061 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4063 PH/02 Patch to exigrep to allow it to work on syslog lines.
4065 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4066 fread() to skip over the body file's header line, because in Cygwin the
4067 header line is locked and is inaccessible.
4069 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4070 co-exist for some time) to make it clear that it is the Exim ID that is
4071 referenced, not the Message-ID: header line.
4073 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4074 string_format() function, because snprintf() does not exist on all
4077 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4078 ${stat: expansion item.
4080 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4081 protocol synchronization error", to keep the pedants happy.
4083 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4084 well as for IRIX systems, when gcc is being used. See the host.c source
4087 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4089 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4092 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4093 transport and to the smtp transport in LMTP mode.
4095 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4097 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4098 run a filter in a subprocess. This could lead to confusion in subsequent
4099 lookups in the parent process. There should also be a search_tidyup() at
4100 the end of the subprocess.
4102 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4103 only if the host matched helo_try_verify_hosts, which caused the
4104 verification to occur when the EHLO/HELO command was issued. The ACL just
4105 tested the remembered result. Now, if a previous verification attempt has
4106 not happened, "verify = helo" does it there and then.
4108 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4110 TK/03 Fix log output including CR from clamd.
4112 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4113 exist provoked a memory error which could cause a segfault.
4115 PH/15 Installed PCRE 6.2
4117 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4119 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4120 of the problem. Specifically, suggested +O2 rather than +O1 for the
4123 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4125 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4126 record, the port was not correctly logged when the outgoing_port log
4127 selector was set (it logged the transort's default port).
4129 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4130 fallback_hosts, and "hosts" in the smtp transport.
4132 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4133 host errors such as "Connection refused".
4135 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4136 authentication with radiusclient 0.4.9:
4138 - Error returned from rc_read_config was caught wrongly
4139 - Username/password not passed on to radius server due to wrong length.
4141 The presumption is that some radiusclient API changes for 4.51/PH/17
4142 were not taken care of correctly. The code is still untested by me (my
4143 Linux distribution still has 0.3.2 of radiusclient), but it was
4144 contributed by a Radius user.
4146 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4147 expanding the "port" option of the smtp transport.
4149 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4150 while reading a MIME header. Thanks to Tom Hughes for a patch.
4152 PH/24 Include config.h inside local_scan.h so that configuration settings are
4155 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4156 that in an ACL for RCPT (for example), you can examine exactly what was
4159 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4160 commands, but it was not correctly comparing the address with the actual
4161 client host address. Thus, it would show the EHLO address in Received:
4162 header lines when this was not necessary.
4164 PH/27 Added the % operator to ${eval:}.
4166 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4167 it should be ignoring failures (because with -C, for example, it has lost
4168 privilege). It wasn't ignoring creation failures other than "already
4171 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4174 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4175 because an input line was too long, either on its own, or by virtue of
4176 too many continuations, the temporary file was not being removed, and the
4177 return code was incorrect.
4179 PH/31 Missing "BOOL" in function definition in filtertest.c.
4181 PH/32 Applied Sieve patches from the maintainer.
4183 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4185 PH/33 Added "verify = not_blind".
4187 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4188 Local/Makefile (with some defaults set). These are used in built scripts
4189 such as exicyclog, but they have never been used in the exim_install
4190 script (though there are many overriding facilities there). I have
4191 arranged that the exim_install script now takes note of these two
4194 PH/35 Installed configuration files for Dragonfly.
4196 PH/36 When a locally submitted message by a trusted user did not contain a
4197 From: header, and the sender address was obtained from -f or from an SMTP
4198 MAIL command, and the trusted user did not use -F to supply a sender
4199 name, $originator_name was incorrectly used when constructing a From:
4200 header. Furthermore, $originator_name was used for submission mode
4201 messages from external hosts without From: headers in a similar way,
4202 which is clearly wrong.
4204 PH/37 Added control=suppress_local_fixups.
4206 PH/38 When log_selector = +received_sender was set, and the addition of the
4207 sender made the log line's construction buffer exactly full, or one byte
4208 less than full, an overflow happened when the terminating "\n" was
4211 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4212 when the result of a list match is failure because a DNS lookup failed.
4214 PH/40 RM_COMMAND is now used in the building process.
4216 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4217 the "build-* directories that it finds.
4219 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4220 address in a domain literal was a prefix of an interface address.
4222 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4223 when verifying a sender address, unless rewrite_headers is false.
4225 PH/44 Wrote a long comment about why errors_to addresses are verified as
4226 recipients, not senders.
4228 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4229 the ratelimit ACL was added.
4231 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4233 PH/46 Added extra information about PostgreSQL errors to the error string.
4235 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4236 avoiding the need to install special zones in a real server. This is
4237 backwards compatible; if it can't find the fake resolver, it drops back.
4238 Thus, both old and new test suites can be run.
4240 TF/02 Added util/ratelimit.pl
4242 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4243 clock is set back in time.
4245 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4246 Candler <B.Candler@pobox.com>.
4248 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4249 OK for addresses that are the result of redirections.
4251 PH/48 A number of further additions for the benefit of the new test suite,
4252 including a fake gethostbyname() that interfaces to the fake DNS resolver
4255 TF/06 The fix for widen_domains has also been applied to qualify_single and
4256 search_parents which are the other dnslookup options that can cause
4259 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4262 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4264 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4265 service" if it encountered an MX record with an empty target hostname.
4266 The message is now "an MX or SRV record indicated no SMTP service".
4268 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4269 if the DNS of the sending site is misconfigured. This is quite a
4270 common situation. This change restores the behaviour of treating a
4271 helo verification defer as a failure.
4273 PH/51 If self=fail was set on a router, the bounce message did not include the
4274 actual error message.
4280 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4282 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4283 command itself ended in error, the underlying message about the transport
4284 filter timeout was being overwritten with the pipe command error. Now the
4285 underlying error message should be appended to the second error message.
4287 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4289 PH/02 Reduce the amount of output that "make" produces by default. Full output
4290 can still be requested.
4292 PH/03 The warning log line about a condition test deferring for a "warn" verb
4293 was being output only once per connection, rather than after each
4294 occurrence (because it was using the same function as for successful
4295 "warn" verbs). This seems wrong, so I have changed it.
4297 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4298 it should not have, which might have caused a crash in the right
4299 circumstances, but probably never did.
4301 PH/04 Installed a modified version of Tony Finch's patch to make submission
4302 mode fix the return path as well as the Sender: header line, and to
4303 add a /name= option so that you can make the user's friendly name appear
4306 TF/03 Added the control = fakedefer ACL modifier.
4308 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4309 Mark Lowes for thorough testing.
4311 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4313 TK/03 Merged latest SRS patch from Miles Wilton.
4315 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4316 with the definition in sysexits.h (which is #included earlier).
4317 Fortunately, Exim does not actually use EX_OK. The code used to try to
4318 preserve the sysexits.h value, by assuming that macro definitions were
4319 scanned for macro replacements. I have been disabused of this notion,
4320 so now the code just undefines EX_OK before #including unistd.h.
4322 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4323 in the smtp transport. When a block could not be written in a single
4324 write() function, the timeout was being re-applied to each part-write.
4325 This seems wrong - if the receiver was accepting one byte at a time it
4326 would take for ever. The timeout is now adjusted when this happens. It
4327 doesn't have to be particularly precise.
4329 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4330 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4332 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4333 without a domain if the check to <postmaster@domain> fails.
4335 SC/01 Eximstats: added -xls and the ability to specify output files
4336 (patch written by Frank Heydlauf).
4338 SC/02 Eximstats: use FileHandles for outputting results.
4340 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4342 SC/04 Eximstats: fixed display of large numbers with -nvr option
4344 SC/05 Eximstats: fixed merging of reports with empty tables.
4346 SC/06 Eximstats: added the -include_original_destination flag
4348 SC/07 Eximstats: removed tabs and trailing whitespace.
4350 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4352 TK/06 MBOX spool code: Add real "From " MBOX separator line
4353 so the .eml file is really in mbox format (even though
4354 most programs do not really care). Patch from Alex Miller.
4356 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4357 The latter is generated from $received_to and is only set if the
4358 message has one envelope recipient. SA can use these headers,
4359 obviously out-of-the-box. Patch from Alex Miller.
4361 PH/08 The ${def test on a variable was returning false if the variable's
4362 value was "0", contrary to what the specification has always said!
4363 The result should be true unless the variable is empty.
4365 PH/09 The syntax error of a character other than { following "${if
4366 def:variable_name" (after optional whitespace) was not being diagnosed.
4367 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4368 accidental colon was present, for example, could give incorrect results.
4370 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4371 result is used (not including appendfile, where other changes are about
4374 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4375 This involved changing a lot of size variables from int to off_t. It
4376 should work with maildirs and everything.
4378 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4379 spamd dying while we are connected to it.
4381 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4384 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4385 function for BDB 4.3.
4387 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4389 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4390 the sources for PCRE have been re-arranged and more files are now
4393 PH/15 The code I had for printing potentially long long variables in PH/11
4394 above was not the best (it lost precision). The length of off_t variables
4395 is now inspected at build time, and an appropriate printing format (%ld
4396 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4397 to be "long long int" or "long int". This is needed for the internal
4398 formatting function string_vformat().
4400 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4401 the configuration file to be ":syslog", then the script "guesses" where
4402 the logs files are, rather than using the compiled in default. In our
4403 case the guess is not the same as the compiled default, so the script
4404 suddenly stopped working when I started to use syslog. The patch checks
4405 to see if log_file_path is "". If so, it attempts to read it from exim
4406 with no configuration file to get the compiled in version, before it
4407 falls back to the previous guessing code."
4409 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4410 implementing BATV in an Exim configuration. See NewStuff for the gory
4413 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4414 Makefile that are specific to HP-UX.
4416 PH/18 If the "use_postmaster" option was set for a recipient callout together
4417 with the "random" option, the postmaster address was used as the MAIL
4418 FROM address for the random test, but not for the subsequent recipient
4419 test. It is now used for both.
4421 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4422 patch removes a few documentation additions to RFC 3028, because the
4423 latest draft now contains them. It adds the new en;ascii-case comparator
4424 and a new error check for 8bit text in MIME parts. Comparator and
4425 require names are now matched exactly. I enabled the subaddress
4426 extension, but it is not well tested yet (read: it works for me)."
4428 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4429 rework some of the code of TK/09 above to avoid the hardwired use of
4430 "%lld" and "long long". Replaced the call to snprintf() with a call to
4433 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4434 records point to non-existent hosts", "retry timeout exceeded", and
4435 "retry time not reached for any host after a long failure period".
4437 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4438 experimental DomainKeys support:
4440 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4441 (2) On an error such as an illegally used "control", the wrong name for
4442 the control was given.
4444 These problems did NOT occur unless DomainKeys support was compiled.
4446 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4448 PH/24 Added ${if match_ip condition.
4450 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4451 databases so that it will be absolutely obvious if a crash occurs in the
4452 DB library. This is a regular occurrence (often caused by mis-matched
4455 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4456 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4457 picked up on a user's system that detects such things. There doesn't seem
4458 to be a gcc warning option for this - only an attribute that has to be
4459 put on the function's prototype. It seems that in Fedora Core 4 they have
4460 set this on a number of new functions. No doubt there will be more in due
4463 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4464 specify a transport. However, if an address that was verified by such a
4465 router was the subject of a callout, Exim crashed because it tried to
4466 read the rcpt_include_affixes from the non-existent transport. Now it
4467 just assumes that the setting of that option is false. This bug was
4468 introduced by 4.51/PH/31.
4470 PH/28 Changed -d+all to exclude +memory, because that information is very
4471 rarely of interest, but it makes the output a lot bigger. People tend to
4472 do -d+all out of habit.
4474 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4475 code in os-type was giving problems when libc.so lives in lib64, like on
4478 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4479 aren't the modern standard, and it seems that some systems' include files
4480 don't always have them. Exim was already checking for some of the newer
4481 ones like T_AAAA, and defining it itself. I've added checks for all the
4482 record types that Exim uses.
4484 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4485 not automatically generating a new one, as it is supposed to. This
4486 prevented TLS from working. If the file did exist, but contained invalid
4487 data, a new version was generated, as expected. It was only the case of a
4488 non-existent file that was broken.
4490 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4491 with a change in libdomainkeys > 0.64.
4493 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4494 from DNS. If the selector record carries the flag, it now has
4495 precedence over the domain-wide flag.
4497 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4499 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4500 the use of a port name that isn't defined in /etc/services) occurred, the
4501 message was deferred as in a normal delivery, and thus remained on the
4502 spool, instead of being failed because of the mua_wrapper setting. This
4503 is now fixed, and I tidied up some of the mua_wrapper messages at the
4506 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4507 the messages in a hash of arrays rather than using individual hashes.
4508 This is a bit cleaner and results in dramatic memory savings, albeit
4509 at a slight CPU cost.
4511 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4512 as requested by Marc Sherman.
4514 SC/10 Eximstats: added histograms for user specified patterns as requested
4517 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4519 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4520 fopen() in the content-scanning modules that did not already have it.
4526 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4527 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4529 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4531 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4533 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4534 to test Sieve filters that use "vacation".
4536 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4537 that changes the way the GnuTLS parameters are stored in the cache file.
4538 The new format can be generated externally. For backward compatibility,
4539 if the data in the cache doesn't make sense, Exim assumes it has read an
4540 old-format file, and it generates new data and writes a new file. This
4541 means that you can't go back to an older release without removing the
4544 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4545 work if there are any delivery delays because "one_time" forces the
4546 parent to be marked "delivered", so its unseen clone is never tried
4547 again. For this reason, Exim now forbids the simultaneous setting of
4550 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4551 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4552 entry if you want to know the details. The fix, however, neglected to
4553 consider the case when local delivery batching is involved. The test for
4554 "previously delivered" was not happening when checking to see if an
4555 address could be batched with a previous (undelivered) one; under
4556 certain circumstances this could lead to multiple deliveries to the same
4559 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4560 in its include files, and this causes problems building Exim.
4562 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4563 header_syntax) but Exim was just ignoring anything given after a slash.
4564 In particular, this caused confusion with an attempt to use "verify =
4565 reverse_host_lookup/defer_ok". An error is now given when options are
4566 supplied for verify items that do not have them. (Maybe reverse_host_
4567 lookup should have a defer_ok option, but that's a different point.)
4569 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4570 defined by RFC 821) to 2048, because there were problems with some AUTH
4571 commands, and RFC 1869 says the size should be increased for extended
4572 SMTP commands that take arguments.
4574 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4577 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4578 "unknown" error; now it says that the functionality isn't in the binary.
4580 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4581 an address' error message when a string expansion fails (syntax or
4582 whatever). Otherwise the password may appear in the log. Following change
4583 PH/42 below, there is no longer a chance of it appearing in a bounce
4586 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4588 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4589 of its addresses was ever tried. (Bugzilla bug #2.)
4591 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4592 the result incorrectly in the debug output. (It correctly added a newline
4593 to what was transported.)
4595 TF/01 Added $received_time.
4597 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4598 commented out examples of how to interface to a virus scanner and to
4599 SpamAssassin. Also added commented examples of av_scanner and
4600 spamd_address settings.
4602 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4603 and controls are allowed in which ACLs. There were a couple of minor
4604 errors. Some of the entries in the conditions table (which is a table of
4605 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4606 negation of where the condition IS allowed.
4608 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4610 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4611 header file does not have a version number, so I've had to invent a new
4612 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4613 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4614 radiusclient), but it was contributed by a Radius user.
4616 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4617 files or directories, for OpenSSL.
4619 PH/19 When an Exim process that is running as root has to create an Exim log
4620 file, it does so in a subprocess that runs as exim:exim so as to get the
4621 ownership right at creation (otherwise, other Exim processes might see
4622 the file with the wrong ownership). There was no test for failure of this
4623 fork() call, which would lead to the process getting stuck as it waited
4624 for a non-existent subprocess. Forks do occasionally fail when resources
4625 run out. I reviewed all the other calls to fork(); they all seem to check
4628 PH/20 When checking for unexpected SMTP input at connect time (before writing
4629 the banner), Exim was not dealing correctly with a non-positive return
4630 from the read() function. If the client had disconnected by this time,
4631 the result was a log entry for a synchronization error with an empty
4632 string after "input=" when read() returned zero. If read() returned -1
4633 (an event I could not check), uninitialized data bytes were printed.
4634 There were reports of junk text (parts of files, etc) appearing after
4637 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4639 PH/22 Added support for macro redefinition, and (re)definition in between
4640 driver and ACL definitions.
4642 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4643 forgetting to use the resulting value; it was using the unexpanded value.
4645 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4646 hadn't been configured. The fix is from Juergen Kreileder, who
4647 understands it better than I do:
4649 "Here's what I see happening with three configured cyrus_sasl
4650 authenticators configured (plain, login, cram-md5):
4652 On startup auth_cyrus_sasl_init() gets called for each of these.
4653 This means three calls to sasl_listmech() without a specified mech_list.
4654 => SASL tests which mechs of all available mechs actually work
4655 => three warnings about OTP not working
4656 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4658 With the patch, sasl_listmech() also gets called three times. But now
4659 SASL's mech_list option is set to the server_mech specified in the the
4660 authenticator. Or in other words, the answer from sasl_listmech()
4661 gets limited to just the mech you're testing for (which is different
4663 => the return list contains just 'plain' or 'login', 'cram-md5' or
4664 nothing depending on the value of ob->server_mech.
4666 I've just tested the patch: Authentication still works fine,
4667 unavailable mechs specified in the exim configuration are still
4668 caught, and the auth.log warnings about OTP are gone."
4670 PH/25 When debugging is enabled, the contents of the command line are added
4671 to the debugging output, even when log_selector=+arguments is not
4674 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4675 answer is "GNU", and only if the return is "GNU/something" is the answer
4678 PH/27 $acl_verify_message is now set immediately after the failure of a
4679 verification in an ACL, and so is available in subsequent modifiers. In
4680 particular, the message can be preserved by coding like this:
4682 warn !verify = sender
4683 set acl_m0 = $acl_verify_message
4685 Previously, $acl_verify_message was set only while expanding "message"
4686 and "log_message" when a very denied access.
4688 PH/28 Modified OS/os.c-Linux with
4690 -#ifndef OS_LOAD_AVERAGE
4691 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4693 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4694 nomenclature these days.)
4696 PH/29 Installed patch from the Sieve maintainer that adds the options
4697 sieve_useraddress and sieve_subaddress to the redirect router.
4699 PH/30 In these circumstances:
4700 . Two addresses routed to the same list of hosts;
4701 . First host does not offer TLS;
4702 . First host accepts first address;
4703 . First host gives temporary error to second address;
4704 . Second host offers TLS and a TLS session is established;
4705 . Second host accepts second address.
4706 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4707 and peerdn, if requested) that were in fact used only for the second
4710 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4711 attention to any local part prefix or suffix that was matched by the
4712 router that accepted the address. It now behaves in the same way as it
4713 does for delivery: the affixes are removed from the local part unless
4714 rcpt_include_affixes is set on the transport.
4716 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4717 timeout during the DATA phase of an incoming message.
4719 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4720 applied a patch sanctioned by the Sieve maintainer.
4722 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4723 the uid or gid is negative. A case of a negative gid caused this to be
4724 noticed. The fix allows for either to be negative.
4726 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4727 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4730 PH/36 The change PH/12 above was broken. Fixed it.
4732 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4733 the grounds that routing the same address twice would always produce the
4734 same answer. This might have been true once, but it is certainly no
4735 longer true now. Routing a child address may depend on the previous
4736 routing that produced that child. Some complicated redirection strategies
4737 went wrong when messages had multiple recipients, and made Exim's
4738 behaviour dependent on the order in which the addresses were given.
4740 I have moved the duplicate checking until after the routing is complete.
4741 Exim scans the addresses that are assigned to local and remote
4742 transports, and removes any duplicates. This means that more work will be
4743 done, as duplicates will always all be routed, but duplicates are
4744 presumably rare, so I don't expect this is of any significance.
4746 For deliveries to pipes, files, and autoreplies, the duplicate checking
4747 still happens during the routing process, since they are not going to be
4750 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4751 It corrects a timeout issue with spamd. This is Ian's comment: "The
4752 background is that sometimes spamd either never reads data from a
4753 connection it has accepted, or it never writes response data. The exiscan
4754 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4755 blindly assumes that writes won't block so it may never time out."
4757 PH/39 Allow G after quota size as well as K and M.
4759 PH/40 The value set for $authenticated_id in an authenticator may not contain
4760 binary zeroes or newlines because the value is written to log lines and
4761 to spool files. There was no check on this. Now the value is run through
4762 the string_printing() function so that such characters are converted to
4763 printable escape sequences.
4765 PH/41 $message_linecount is a new variable that contains the total number of
4766 lines in the message. Compare $body_linecount, which is the count for the
4769 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4770 bounce and delay warning messages, except in certain special cases, which
4773 (a) An SMTP error message from a remote host;
4774 (b) A message specified in a :fail: redirection;
4775 (c) A message specified in a "fail" command in a system filter;
4776 (d) A message specified in a FAIL return from the queryprogram router;
4777 (e) A message specified by the cannot_route_message router option.
4779 In these cases only, Exim does include the error details in bounce and
4780 warning messages. There are also a few cases where bland messages such
4781 as "unrouteable address" or "local delivery error" are given.
4783 PH/43 $value is now also set for the "else" part of a ${run expansion.
4785 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4786 being worked on, but at least Exim now implements the latest version to
4789 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4790 process to complete was treated as a delivery failure, a timeout while
4791 writing the message to the pipe was logged, but erroneously treated as a
4792 successful delivery. Such timeouts include transport filter timeouts. For
4793 consistency with the overall process timeout, these timeouts are now
4794 treated as errors, giving rise to delivery failures by default. However,
4795 there is now a new Boolean option for the pipe transport called
4796 timeout_defer, which, if set TRUE, converts the failures into defers for
4797 both kinds of timeout. A transport filter timeout is now identified in
4800 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4801 systems where "make" and "gmake" are different, calling "gmake" at top
4802 level broke things. I've arranged for the value of $(MAKE) to be passed
4803 from the Makefile to this script so that it can call the same version of
4807 A note about Exim versions 4.44 and 4.50
4808 ----------------------------------------
4810 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4811 changes of various kinds. As a consequence, a big documentation update was
4812 needed. This delayed the release for rather longer than seemed good, especially
4813 in the light of a couple of (minor) security issues. Therefore, the changes
4814 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4815 release. So 4.44 and 4.50 are in effect two different branches that both start
4818 I have left the 4.50 change log unchanged; it contains all the changes since
4819 4.43. The change log for 4.44 is below; many of its items are identical to
4820 those for 4.50. This seems to be the most sensible way to preserve the
4821 historical information.
4827 1. Minor wording change to the doc/README.SIEVE file.
4829 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4830 computation of the current number of files was incorrect.
4832 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4833 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4836 4. Give more explanation in the error message when the command for a transport
4837 filter fails to execute.
4839 5. There are several places where Exim runs a non-Exim command in a
4840 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4841 was being done only for the command run by the queryprogram router. It is
4842 now done for all such subprocesses. The other cases are: ${run, transport
4843 filters, and the commands run by the lmtp and pipe transports.
4845 6. Added CONFIGURE_GROUP build-time option.
4847 7. Some older OS have a limit of 256 on the maximum number of file
4848 descriptors. Exim was using setrlimit() to set 1000 as a large value
4849 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4850 systems. I've change it so that if it can't get 1000, it tries for 256.
4852 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4853 was an oversight, and furthermore, ever since the addition of extra
4854 controls (e.g. 4.43/32), the checks on when to allow different forms of
4855 "control" were broken. There should now be diagnostics for all cases when a
4856 control that does not make sense is encountered.
4858 9. Added the /retain_sender option to "control=submission".
4860 10. $recipients is now available in the predata ACL (oversight).
4862 11. Tidy the search cache before the fork to do a delivery from a message
4863 received from the command line. Otherwise the child will trigger a lookup
4864 failure and thereby defer the delivery if it tries to use (for example) a
4865 cached ldap connection that the parent has called unbind on.
4867 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4868 of $address_data from the recipient verification was clobbered by the
4869 sender verification.
4871 13. The value of address_data from a sender verification is now available in
4872 $sender_address_data in subsequent conditions in the ACL statement.
4874 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4876 15. Added a new option "connect=<time>" to callout options, to set a different
4879 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4880 was its contents. (It was OK if the option was not defined at all.)
4882 17. A "Completed" log line is now written for messages that are removed from
4883 the spool by the -Mrm option.
4885 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4886 information about exactly what failed.
4888 19. Added -dd to debug only the daemon process.
4890 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4891 handles timeouts, both on the server side and network timeouts. Renamed the
4892 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4894 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4895 It is now set to "smtps".
4897 22. $host_address is now set to the target address during the checking of
4898 ignore_target_hosts.
4900 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4901 being passed; this would have caused $sender_host_name to have been used if
4902 matching the list had actually called for a host name (not very likely,
4903 since this list is usually IP addresses). A host name is now passed as
4906 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4907 code that specifies a non-restarting handler (typically sigaction() in
4908 modern systems) in an attempt to fix a rare and obscure crash bug.
4910 25. Narrowed the window for a race in the daemon that could cause it to ignore
4911 SIGCHLD signals. This is not a major problem, because they are used only to
4912 wake it up if nothing else does.
4914 26. A malformed maildirsize file could cause Exim to calculate negative values
4915 for the mailbox size or file count. Odd effects could occur as a result.
4916 The maildirsize information is now recalculated if the size or filecount
4919 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4920 support for a long time. Removed HAVE_SYS_VFS_H.
4922 28. Installed the latest version of exipick from John Jetmore.
4924 29. In an address list, if the pattern was not a regular expression, an empty
4925 subject address (from a bounce message) matched only if the pattern was an
4926 empty string. Non-empty patterns were not even tested. This was the wrong
4927 because it is perfectly reasonable to use an empty address as part of a
4928 database query. An empty address is now tested by patterns that are
4929 lookups. However, all the other forms of pattern expect the subject to
4930 contain a local part and a domain, and therefore, for them, an empty
4931 address still always fails if the pattern is not itself empty.
4933 30. Exim went into a mad DNS loop when attempting to do a callout where the
4934 host was specified on an smtp transport, and looking it up yielded more
4935 than one IP address.
4937 31. Re-factored the code for checking spool and log partition space into a
4938 function that finds that data and another that does the check. The former
4939 is then used to implement four new variables: $spool_space, $log_space,
4940 $spool_inodes, and $log_inodes.
4942 32. The RFC2047 encoding function was originally intended for short strings
4943 such as real names; it was not keeping to the 75-character limit for
4944 encoded words that the RFC imposes. It now respects the limit, and
4945 generates multiple encoded words if necessary. To be on the safe side, I
4946 have increased the buffer size for the ${rfc2047: expansion operator from
4949 33. It is now permitted to omit both strings after an "if" condition; if the
4950 condition is true, the result is "true". As before, when the second string
4951 is omitted, a false condition yields an empty string. This makes it less
4952 cumbersome to write custom ACL and router conditions.
4954 34. Failure to deliver a bounce message always caused it to be frozen, even if
4955 there was an errors_to setting on the router. The errors_to setting is now
4958 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4959 canonical form (fully expanded) before being placed in
4960 $sender_host_address.
4962 36. The table in the code that translates DNS record types into text (T_A to
4963 "A" for instance) was missing entries for NS and CNAME. It is just possible
4964 that this could have caused confusion if both these types were looked up
4965 for the same domain, because the text type is used as part of Exim's
4966 per-process caching. But the chance of anyone hitting this buglet seems
4969 37. The dnsdb lookup has been extended in a number of ways.
4971 (1) There is a new type, "zns", which walks up the domain tree until it
4972 finds some nameserver records. It should be used with care.
4974 (2) There is a new type, "mxh", which is like "mx" except that it returns
4975 just the host names, not the priorities.
4977 (3) It is now possible to give a list of domains (or IP addresses) to be
4978 looked up. The behaviour when one of the lookups defers can be
4979 controlled by a keyword.
4981 (4) It is now possible to specify the separator character for use when
4982 multiple records are returned.
4984 38. The dnslists ACL condition has been extended: it is now possible to supply
4985 a list of IP addresses and/or domains to be looked up in a particular DNS
4988 39. Added log_selector=+queue_time_overall.
4990 40. When running the queue in the test harness, wait just a tad after forking a
4991 delivery process, to get repeatability of debugging output.
4993 41. Include certificate and key file names in error message when GnuTLS fails
4994 to set them up, because the GnuTLS error message doesn't include the name
4995 of the failing file when there is a problem reading it.
4997 42. Allow both -bf and -bF in the same test run.
4999 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5001 44. The "Exiscan patch" is now merged into the mainline Exim source.
5003 45. Sometimes the final signoff response after QUIT could fail to get
5004 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5005 < 0 before doing a fflush(). This bug looks as though it goes back to the
5006 introduction of TLS in release 3.20, but "sometimes" must have been rare
5007 because the tests only now provoked it.
5009 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5010 (this can affect the format of dates).
5012 47. exim_tidydb, when checking for the continued existence of a message for
5013 which it has found a message-specific retry record, was not finding
5014 messages that were in split spool directories. Consequently, it was
5015 deleting retry records that should have stayed in existence.
5017 48. Steve fixed some bugs in eximstats.
5019 49. The SPA authentication driver was not abandoning authentication and moving
5020 on to the next authenticator when an expansion was forced to fail,
5021 contradicting the general specification for all authenticators. Instead it
5022 was generating a temporary error. It now behaves as specified.
5024 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5025 (the order specifies the preference for clients). The order is now AES256,
5026 AES128, 3DES, ARCFOUR128.
5028 51. Small patch to Sieve code - explicitly set From: when generating an
5031 52. Exim crashed if a remote delivery caused a very long error message to be
5032 recorded - for instance if somebody sent an entire SpamAssassin report back
5033 as a large number of 550 error lines. This bug was coincidentally fixed by
5034 increasing the size of one of Exim's internal buffers (big_buffer) that
5035 happened as part of the Exiscan merge. However, to be on the safe side, I
5036 have made the code more robust (and fixed the comments that describe what
5039 53. Now that there can be additional text after "Completed" in log lines (if
5040 the queue_time_overall log selector is set), a one-byte patch to exigrep
5041 was needed to allow it to recognize "Completed" as not the last thing in
5044 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5045 patch that reportedly fixes this has been added. I am not expert enough to
5046 create a test for it. This is what the patch creator wrote:
5048 "I found a little strange behaviour of ldap code when working with
5049 Windows 2003 AD Domain, where users was placed in more than one
5050 Organization Units. When I tried to give exim partial DN, the exit code
5051 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5052 But simultaneously result of request was absolutely normal ldap result,
5053 so I produce this patch..."
5055 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5056 REFERENCE, so I have modified the code to exclude the patch when that macro
5059 55. Some experimental protocols are using DNS PTR records for new purposes. The
5060 keys for these records are domain names, not reversed IP addresses. The
5061 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5062 leaves it alone. Component reversal etc. now happens only for IP addresses.
5065 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5067 57. Double the size of the debug message buffer (to 2048) so that more of very
5068 long debug lines gets shown.
5070 58. The exicyclog utility now does better if the number of log files to keep
5071 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5073 59. Two changes related to the smtp_active_hostname option:
5075 (1) $smtp_active_hostname is now available as a variable.
5076 (2) The default for smtp_banner uses $smtp_active_hostname instead
5077 of $primary_hostname.
5079 60. The host_aton() function is supposed to be passed a string that is known
5080 to be a valid IP address. However, in the case of IPv6 addresses, it was
5081 not checking this. This is a hostage to fortune. Exim now panics and dies
5082 if the condition is not met. A case was found where this could be provoked
5083 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5084 components; fortuitously, this particular loophole had already been fixed
5085 by change 4.50/55 above.
5087 If there are any other similar loopholes, the new check in host_aton()
5088 itself should stop them being exploited. The report I received stated that
5089 data on the command line could provoke the exploit when Exim was running as
5090 exim, but did not say which command line option was involved. All I could
5091 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5092 running as the user.
5095 61. There was a buffer overflow vulnerability in the SPA authentication code
5096 (which came originally from the Samba project). I have added a test to the
5097 spa_base64_to_bits() function which I hope fixes it.
5100 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5101 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5103 63. The daemon start-up calls getloadavg() while still root for those OS that
5104 need the first call to be done as root, but it missed one case: when
5105 deliver_queue_load_max is set with deliver_drop_privilege. This is
5106 necessary for the benefit of the queue runner, because there is no re-exec
5107 when deliver_drop_privilege is set.
5109 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5110 This has been fixed.
5112 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5113 was in use, was not putting the data itself into the right store pool;
5114 consequently, it could be overwritten for a subsequent message in the same
5115 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5118 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5120 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5121 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5122 the function were treating the return as a boolean value, which happened to
5123 work because 0=false and not-0=true, but is not correct code.
5125 68. The host_aton() function was not handling scoped IPv6 addresses (those
5126 with, for example, "%eth0" on the end) correctly.
5128 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5129 negated items (that is, ~something) in unsigned ints. Some compilers
5130 apparently mutter when there is no cast.
5132 70. If an address verification called from an ACL failed, and did not produce a
5133 user-specific message (i.e. there was only a "system" message), nothing was
5134 put in $acl_verify_message. In this situation, it now puts the system
5137 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5138 session; change 4.31/43 added the unwanted input to the log line - except
5139 that it did not do this in the start of session case. It now does.
5141 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5142 This is wrong and can cause the other end to generate a synchronization
5143 error if it is another Exim or anything else that does the synchronization
5144 check. A QUIT command is no longer sent after a timeout.
5146 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5147 during host lookups.
5149 74. The defer_ok option of callout verification was not working if it was used
5150 when verifying addresses in header lines, that is, for this case:
5152 verify = header_sender/callout=defer_ok
5154 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5155 those file descriptors could be used for SMTP connections. If anything
5156 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5157 could be sent to the SMTP client, causing chaos. The daemon now opens
5158 stdin, stdout, and stderr to /dev/null when it puts itself into the
5161 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5162 log by default. The user can override this with suitable Perl magic.
5164 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5165 the log message when discard triggers, was not working for the DATA ACL or
5166 for the non-SMTP ACL.
5168 78. Error message wording change in sieve.c.
5170 79. If smtp_accept_max_per_host was set, the number of connections could be
5171 restricted to fewer than expected, because the daemon was trying to set up
5172 a new connection before checking whether the processes handling previous
5173 connections had finished. The check for completed processes is now done
5174 earlier. On busy systems, this bug wouldn't be noticed because something
5175 else would have woken the daemon, and it would have reaped the completed
5178 80. If a message was submitted locally by a user whose login name contained one
5179 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5180 It caused a spool format error. I have fixed the spool reading code. A
5181 related problem was that the "from" clause in the Received: line became
5182 illegal because of the space(s). It is now covered by ${quote_local_part.
5184 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5187 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5189 83. Patch from Sieve maintainer to fix unterminated string problem in
5190 "vacation" handling.
5192 84. Some minor changes to the Linux configuration files to help with other
5193 OS variants using glibc.
5195 85. One more patch for Sieve to update vacation handling to latest spec.
5198 ----------------------------------------------------
5199 See the note above about the 4.44 and 4.50 releases.
5200 ----------------------------------------------------
5206 1. Change 4.43/35 introduced a bug that caused file counts to be
5207 incorrectly computed when quota_filecount was set in an appendfile
5210 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5211 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5214 3. Give more explanation in the error message when the command for a transport
5215 filter fails to execute.
5217 4. There are several places where Exim runs a non-Exim command in a
5218 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5219 was being done only for the command run by the queryprogram router. It is
5220 now done for all such subprocesses. The other cases are: ${run, transport
5221 filters, and the commands run by the lmtp and pipe transports.
5223 5. Some older OS have a limit of 256 on the maximum number of file
5224 descriptors. Exim was using setrlimit() to set 1000 as a large value
5225 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5226 systems. I've change it so that if it can't get 1000, it tries for 256.
5228 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5229 was an oversight, and furthermore, ever since the addition of extra
5230 controls (e.g. 4.43/32), the checks on when to allow different forms of
5231 "control" were broken. There should now be diagnostics for all cases when a
5232 control that does not make sense is encountered.
5234 7. $recipients is now available in the predata ACL (oversight).
5236 8. Tidy the search cache before the fork to do a delivery from a message
5237 received from the command line. Otherwise the child will trigger a lookup
5238 failure and thereby defer the delivery if it tries to use (for example) a
5239 cached ldap connection that the parent has called unbind on.
5241 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5242 of $address_data from the recipient verification was clobbered by the
5243 sender verification.
5245 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5246 was its contents. (It was OK if the option was not defined at all.)
5248 11. A "Completed" log line is now written for messages that are removed from
5249 the spool by the -Mrm option.
5251 12. $host_address is now set to the target address during the checking of
5252 ignore_target_hosts.
5254 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5255 being passed; this would have caused $sender_host_name to have been used if
5256 matching the list had actually called for a host name (not very likely,
5257 since this list is usually IP addresses). A host name is now passed as
5260 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5261 code that specifies a non-restarting handler (typically sigaction() in
5262 modern systems) in an attempt to fix a rare and obscure crash bug.
5264 15. Narrowed the window for a race in the daemon that could cause it to ignore
5265 SIGCHLD signals. This is not a major problem, because they are used only to
5266 wake it up if nothing else does.
5268 16. A malformed maildirsize file could cause Exim to calculate negative values
5269 for the mailbox size or file count. Odd effects could occur as a result.
5270 The maildirsize information is now recalculated if the size or filecount
5273 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5274 support for a long time. Removed HAVE_SYS_VFS_H.
5276 18. Updated exipick to current release from John Jetmore.
5278 19. Allow an empty sender to be matched against a lookup in an address list.
5279 Previously the only cases considered were a regular expression, or an
5282 20. Exim went into a mad DNS lookup loop when doing a callout where the
5283 host was specified on the transport, if the DNS lookup yielded more than
5286 21. The RFC2047 encoding function was originally intended for short strings
5287 such as real names; it was not keeping to the 75-character limit for
5288 encoded words that the RFC imposes. It now respects the limit, and
5289 generates multiple encoded words if necessary. To be on the safe side, I
5290 have increased the buffer size for the ${rfc2047: expansion operator from
5293 22. Failure to deliver a bounce message always caused it to be frozen, even if
5294 there was an errors_to setting on the router. The errors_to setting is now
5297 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5298 canonical form (fully expanded) before being placed in
5299 $sender_host_address.
5301 24. Updated eximstats to version 1.33
5303 25. Include certificate and key file names in error message when GnuTLS fails
5304 to set them up, because the GnuTLS error message doesn't include the name
5305 of the failing file when there is a problem reading it.
5307 26. Expand error message when OpenSSL has problems setting up cert/key files.
5310 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5311 (this can affect the format of dates).
5313 28. exim_tidydb, when checking for the continued existence of a message for
5314 which it has found a message-specific retry record, was not finding
5315 messages that were in split spool directories. Consequently, it was
5316 deleting retry records that should have stayed in existence.
5318 29. eximstats updated to version 1.35
5319 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5320 1.35 - bugfix such that pie charts by volume are generated correctly
5322 30. The SPA authentication driver was not abandoning authentication and moving
5323 on to the next authenticator when an expansion was forced to fail,
5324 contradicting the general specification for all authenticators. Instead it
5325 was generating a temporary error. It now behaves as specified.
5327 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5328 (the order specifies the preference for clients). The order is now AES256,
5329 AES128, 3DES, ARCFOUR128.
5331 31. Small patch to Sieve code - explicitly set From: when generating an
5334 32. Exim crashed if a remote delivery caused a very long error message to be
5335 recorded - for instance if somebody sent an entire SpamAssassin report back
5336 as a large number of 550 error lines. This bug was coincidentally fixed by
5337 increasing the size of one of Exim's internal buffers (big_buffer) that
5338 happened as part of the Exiscan merge. However, to be on the safe side, I
5339 have made the code more robust (and fixed the comments that describe what
5342 33. Some experimental protocols are using DNS PTR records for new purposes. The
5343 keys for these records are domain names, not reversed IP addresses. The
5344 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5345 leaves it alone. Component reversal etc. now happens only for IP addresses.
5348 34. The host_aton() function is supposed to be passed a string that is known
5349 to be a valid IP address. However, in the case of IPv6 addresses, it was
5350 not checking this. This is a hostage to fortune. Exim now panics and dies
5351 if the condition is not met. A case was found where this could be provoked
5352 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5353 components; fortuitously, this particular loophole had already been fixed
5354 by change 4.50/55 or 4.44/33 above.
5356 If there are any other similar loopholes, the new check in host_aton()
5357 itself should stop them being exploited. The report I received stated that
5358 data on the command line could provoke the exploit when Exim was running as
5359 exim, but did not say which command line option was involved. All I could
5360 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5361 running as the user.
5364 35. There was a buffer overflow vulnerability in the SPA authentication code
5365 (which came originally from the Samba project). I have added a test to the
5366 spa_base64_to_bits() function which I hope fixes it.
5369 36. The daemon start-up calls getloadavg() while still root for those OS that
5370 need the first call to be done as root, but it missed one case: when
5371 deliver_queue_load_max is set with deliver_drop_privilege. This is
5372 necessary for the benefit of the queue runner, because there is no re-exec
5373 when deliver_drop_privilege is set.
5375 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5376 was in use, was not putting the data itself into the right store pool;
5377 consequently, it could be overwritten for a subsequent message in the same
5378 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5381 38. Sometimes the final signoff response after QUIT could fail to get
5382 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5383 < 0 before doing a fflush(). This bug looks as though it goes back to the
5384 introduction of TLS in release 3.20, but "sometimes" must have been rare
5385 because the tests only now provoked it.
5391 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5392 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5393 Now it returns an integer. A number of calls were still expecting a T/F
5394 return. Fortuitously, in all cases, the tests worked in OK situations,
5395 which is the norm. However, things would have gone wrong on any write
5396 failures on the smtp file descriptor. This function is used when sending
5397 messages over SMTP and also when doing verify callouts.
5399 2. When Exim is called to do synchronous delivery of a locally submitted
5400 message (the -odf or -odi options), it no longer closes stderr before doing
5403 3. Implemented the mua_wrapper option.
5405 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5407 5. Implemented the functions header_remove(), header_testname(),
5408 header_add_at_position(), and receive_remove_recipient(), and exported them
5411 6. If an ACL "warn" statement specified the addition of headers, Exim already
5412 inserted X-ACL-Warn: at the start if there was no header name. However, it
5413 was not making this test for the second and subsequent header lines if
5414 there were newlines in the string. This meant that an invalid header could
5415 be inserted if Exim was badly configured.
5417 7. Allow an ACL "warn" statement to add header lines at the start or after all
5418 the Received: headers, as well as at the end.
5420 8. Added the rcpt_4xx retry error code.
5422 9. Added postmaster_mailfrom=xxx to callout verification option.
5424 10. Added mailfrom=xxxx to the callout verification option, for verify=
5427 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5428 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5430 12. Inserted some casts to stop certain compilers warning when using pointer
5431 differences as field lengths or precisions in printf-type calls (mostly
5432 affecting debugging statements).
5434 13. Added optional readline() support for -be (dynamically loaded).
5436 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5437 same clock tick as a message's arrival, so that its received time was the
5438 same as the "first fail" time on the retry record, and that message
5439 remained on the queue past the ultimate address timeout, every queue runner
5440 would try a delivery (because it was past the ultimate address timeout) but
5441 after another failure, the ultimate address timeout, which should have then
5442 bounced the address, did not kick in. This was a "< instead of <=" error;
5443 in most cases the first failure would have been in the next clock tick
5444 after the received time, and all would be well.
5446 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5447 being recognized when the domain list was tested by the match_domain
5448 condition in an expansion string.
5450 16. Added the ${str2b64: operator.
5452 17. Exim was always calling setrlimit() to set a large limit for the number of
5453 processes, without checking whether the existing limit was already
5454 adequate. (It did check for the limit on file descriptors.) Furthermore,
5455 errors from getrlimit() and setrlimit() were being ignored. Now they are
5456 logged to the main and panic logs, but Exim does carry on, to try to do its
5457 job under whatever limits there are.
5459 18. Imported PCRE 5.0.
5461 19. Trivial typo in log message " temporarily refused connection" (the leading
5464 20. If the log selector return_path_on_delivery was set and an address was
5465 redirected to /dev/null, the delivery process crashed because it assumed
5466 that a return path would always be set for a "successful" delivery. In this
5467 case, the whole delivery is bypassed as an optimization, and therefore no
5470 21. Internal re-arrangement: the function for sending a challenge and reading
5471 a response while authentication was assuming a zero-terminated challenge
5472 string. It's now changed to take a pointer and a length, to allow for
5473 binary data in such strings.
5475 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5477 23. Exim was not respecting finduser_retries when seeking the login of the
5478 uid under which it was called; it was always trying 10 times. (The default
5479 setting of finduser_retries is zero.) Also, it was sleeping after the final
5480 failure, which is pointless.
5482 24. Implemented tls_on_connect_ports.
5484 25. Implemented acl_smtp_predata.
5486 26. If the domain in control=submission is set empty, Exim assumes that the
5487 authenticated id is a complete email address when it generates From: or
5488 Sender: header lines.
5490 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5491 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5492 chown and chgrp in /bin and hostname in /usr/bin.
5494 28. Exim was keeping the "process log" file open after each use, just as it
5495 does for the main log. This opens the possibility of it remaining open for
5496 long periods when the USR1 signal hits a daemon. Occasional processlog
5497 errors were reported, that could have been caused by this. Anyway, it seems
5498 much more sensible not to leave this file open at all, so that is what now
5501 29. The long-running daemon process does not normally write to the log once it
5502 has entered its main loop, and it closes the log before doing so. This is
5503 so that log files can straightforwardly be renamed and moved. However,
5504 there are a couple of unusual error situations where the daemon does write
5505 log entries, and I had neglected to close the log afterwards.
5507 30. The text of an SMTP error response that was received during a remote
5508 delivery was being truncated at 512 bytes. This is too short for some of
5509 the long messages that one sometimes sees. I've increased the limit to
5512 31. It is now possible to make retry rules that apply only when a message has a
5513 specific sender, in particular, an empty sender.
5515 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5516 it possible to be selective about when SMTP synchronization is enforced.
5518 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5520 32. Implemented hosts_connection_nolog.
5522 33. Added an ACL for QUIT.
5524 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5527 35. Added mailbox_size and mailbox_filecount to appendfile.
5529 36. Added control = no_multiline_responses to ACLs.
5531 37. There was a bug in the logic of the code that waits for the clock to tick
5532 in the case where the clock went backwards by a substantial amount such
5533 that the microsecond fraction of "now" was more than the microsecond
5534 fraction of "then" (but the whole seconds number was less).
5536 38. Added support for the libradius Radius client library this is found on
5537 FreeBSD (previously only the radiusclient library was supported).
5543 1. When certain lookups returned multiple values in the form name=value, the
5544 quoting of the values was not always being done properly. Specifically:
5545 (a) If the value started with a double quote, but contained no whitespace,
5547 (b) If the value contained whitespace other than a space character (i.e.
5548 tabs or newlines or carriage returns) it was not quoted.
5549 This fix has been applied to the mysql and pgsql lookups by writing a
5550 separate quoting function and calling it from the lookup code. The fix
5551 should probably also be applied to nisplus, ibase and oracle lookups, but
5552 since I cannot test any of those, I have not disturbed their existing code.
5554 2. A hit in the callout cache for a specific address caused a log line with no
5555 reason for rejecting RCPT. Now it says "Previous (cached) callout
5556 verification failure".
5558 3. There was an off-by-one bug in the queryprogram router. An over-long
5559 return line was truncated at 256 instead of 255 characters, thereby
5560 overflowing its buffer with the terminating zero. As well as fixing this, I
5561 have increased the buffer size to 1024 (and made a note to document this).
5563 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5564 when Exim is waiting for an SMTP response from a remote server, Exim
5565 restarts its select() call on the socket, thereby resetting its timeout.
5566 This is not a problem when such interrupts are rare. Somebody set up a cron
5567 job to run exiwhat every 2 minutes, which is less than the normal select()
5568 timeout (5 or 10 minutes). This meant that the select() timeout never
5569 kicked in because it was always reset. I have fixed this by comparing the
5570 time when an interrupt arrives with the time at the start of the first call
5571 to select(). If more time than the timeout has elapsed, the interrupt is
5572 treated as a timeout.
5574 5. Some internal re-factoring in preparation for the addition of Sieve
5575 extensions (by MH). In particular, the "personal" test is moved to a
5576 separate function, and given an option for scanning Cc: and Bcc: (which is
5577 not set for Exim filters).
5579 6. When Exim created an email address using the login of the caller as the
5580 local part (e.g. when creating a From: or Sender: header line), it was not
5581 quoting the local part when it contained special characters such as @.
5583 7. Installed new OpenBSD configuration files.
5585 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5586 try to make them clearer.
5588 9. Callout options, other than the timeout value, were being ignored when
5589 verifying sender addresses in header lines. For example, when using
5591 verify = header_sender/callout=no_cache
5593 the cache was (incorrectly) being used.
5595 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5596 only the Cygwin environment.
5598 11. When return_path_on_delivery was set as a log selector, if different remote
5599 addresses in the same message used different return paths and parallel
5600 remote delivery occurred, the wrong values would sometimes be logged.
5601 (Whenever a remote delivery process finished, the return path value from
5602 the most recently started remote delivery process was logged.)
5604 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5605 lines when AUTH and/or TLS are in use. This is the "received protocol"
5606 field. Exim used to use "asmtp" for authenticated SMTP, without any
5607 indication (in the protocol name) for TLS use. Now it follows the RFC and
5608 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5609 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5610 names appear in log lines as well as in Received: header lines.
5612 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5613 extensions, and comparison tests, and to fix some bugs.
5615 14. Changes to the "personal" filter test:
5617 (1) The test was buggy in that it was just doing the equivalent of
5618 "contains" tests on header lines. For example, if a user's address was
5619 anne@some.where, the "personal" test would incorrectly be true for
5621 To: susanne@some.where
5623 This test is now done by extracting each address from the header in turn,
5624 and checking the entire address. Other tests that are part of "personal"
5625 are now done using regular expressions (for example, to check local parts
5626 of addresses in From: header lines).
5628 (2) The list of non-personal local parts in From: addresses has been
5629 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5630 taken from the Sieve specification recommendations.
5632 (3) If the message contains any header line starting with "List-" it is
5633 treated as non-personal.
5635 (4) The test for "circular" in the Subject: header line has been removed
5636 because it now seems ill-conceived.
5638 15. Minor typos in src/EDITME comments corrected.
5640 16. Installed latest exipick from John Jetmore.
5642 17. If headers_add on a router specified a text string that was too long for
5643 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5644 of string_sprintf() is now avoided.
5646 18. $message_body_size was not set (it was always zero) when running the DATA
5647 ACL and the local_scan() function.
5649 19. For the "mail" command in an Exim filter, no default was being set for
5650 the once_repeat time, causing a random time value to be used if "once" was
5651 specified. (If the value happened to be <= 0, no repeat happened.) The
5652 default is now 0s, meaning "never repeat". The "vacation" command was OK
5653 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5654 (I found it when inspecting the code).
5656 20. There is now an overall timeout for performing a callout verification. It
5657 defaults to 4 times the callout timeout, which applies to individual SMTP
5658 commands during the callout. The overall timeout applies when there is more
5659 than one host that can be tried. The timeout is checked before trying the
5660 next host. This prevents very long delays if there are a large number of
5661 hosts and all are timing out (e.g. when the network connections are timing
5662 out). The value of the overall timeout can be changed by specifying an
5663 additional sub-option for "callout", called "maxwait". For example:
5665 verify = sender/callout=5s,maxwait=20s
5667 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5668 to the end before writing, but this should make it even safer).
5670 22. Exim was forgetting that it had advertised PIPELINING for the second and
5671 subsequent messages on an SMTP connection. It was also not resetting its
5672 memory on STARTTLS and an internal HELO.
5674 23. When Exim logs an SMTP synchronization error within a session, it now
5675 records whether PIPELINING has been advertised or not.
5677 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5678 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5679 rather than long int.
5681 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5683 26. Added the never_mail option to autoreply.
5689 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5690 crash if the getsockname() call failed; this can happen if a connection is
5691 closed very soon after it is established. The problem was simply in the
5692 order in which certain operations were done, causing Exim to try to write
5693 to the SMTP stream before it had set up the file descriptor. The bug has
5694 been fixed by making things happen in the correct order.
5700 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5701 before the connection was closed, thus losing the rejection response.
5703 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5704 some early Solaris releases, but causes trouble in current releases where
5705 socklen_t is defined.
5707 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5710 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5713 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5714 that releasing the top of it at the end releases what was used for sub-
5715 expansions (unless the block got too big). However, discard this block if
5716 the first thing is a variable or header, so that we can use its block when
5717 it is dynamic (useful for very large $message_headers, for example).
5719 6. Lookups now cache *every* query, not just the most recent. A new, separate
5720 store pool is used for this. It can be recovered when all lookup caches are
5721 flushed. Lookups now release memory at the end of their result strings.
5722 This has involved some general refactoring of the lookup sources.
5724 7. Some code has been added to the store_xxx() functions to reduce the amount
5725 of flapping under certain conditions.
5727 8. log_incoming_interface used to affect only the <= reception log lines. Now
5728 it causes the local interface and port to be added to several more SMTP log
5729 lines, for example "SMTP connection from", and rejection lines.
5731 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5733 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5735 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5736 could be overwritten at the end of the current message (or the start of a
5737 new message if it was set in a HELO ACL). The value is now preserved for
5738 the duration of the SMTP connection.
5740 12. If a transport had a headers_rewrite setting, and a matching header line
5741 contained an unqualified address, that address was qualified, even if it
5742 did not match any rewriting rules. The underlying bug was that the values
5743 of the flags that permit the existence of unqualified sender and recipient
5744 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5745 non-local messages, and by -bnq for local messages) were not being
5746 preserved with the message after it was received.
5748 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5749 "next input=" as part of the text comprising the host identity instead of
5750 the correct text. The code was using the same buffer for two different
5751 strings. However, depending on which order the printing function evaluated
5752 its arguments, the bug did not always show up. Under Linux, for example, my
5753 test suite worked just fine.
5755 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5756 doesn't work with some older versions of Perl. It has been changed to "my",
5757 which in any case is probably the better facility to use.
5759 15. A really picky compiler found some instances of statements for creating
5760 error messages that either had too many or two few arguments for the format
5763 16. The size of the buffer for calls to the DNS resolver has been increased
5764 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5765 for addresses that have a lot of PTR records. This alleviates a problem; it
5766 does not fully solve it.
5768 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5769 buffer now truncates the list and logs the incident, which is the same
5770 action as happens when Exim is looking up a host name and its aliases.
5771 Previously in this situation something unpredictable would happen;
5772 sometimes it was "internal error: store_reset failed".
5774 18. If a server dropped the connection unexpectedly when an Exim client was
5775 using GnuTLS and trying to read a response, the client delivery process
5776 crashed while trying to generate an error log message.
5778 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5779 string, for example:
5781 warn message = H1: something\nH2: something
5783 the text was added as a single header line from Exim's point of view
5784 though it ended up OK in the delivered message. However, searching for the
5785 second and subsequent header lines using $h_h2: did not work. This has been
5786 fixed. Similarly, if a system filter added multiple headers in this way,
5787 the routers could not see them.
5789 20. Expanded the error message when iplsearch is called with an invalid key to
5790 suggest using net-iplsearch in a host list.
5792 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5793 ACLs are no longer actually imposed (and a message to that effect is
5796 22. If a "gecos" field in a passwd entry contained escaped characters, in
5797 particular, if it contained a \" sequence, Exim got it wrong when building
5798 a From: or a Sender: header from that name. A second bug also caused
5799 incorrect handling when an unquoted " was present following a character
5800 that needed quoting.
5802 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5803 was not being matched caselessly.
5805 24. Arranged for all hyphens in the exim.8 source to be escaped with
5808 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5809 properly. Recipient callout cache records were still being keyed to include
5810 the sender, even when use_sender was set false. This led to far more
5811 callouts that were necessary. The sender is no longer included in the key
5812 when use_sender is false.
5814 26. Added "control = submission" modifier to ACLs.
5816 27. Added the ${base62d: operator to decode base 62 numbers.
5818 28. dnsdb lookups can now access SRV records.
5820 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5821 the configuration file.
5823 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5824 (-v) mode. This makes the output for a verbose queue run more intelligible.
5826 31. Added a use_postmaster feature to recipient callouts.
5828 32. Added the $body_zerocount variable, containing the number of binary zero
5829 bytes in the message body.
5831 33. The time of last modification of the "new" subdirectory is now used as the
5832 "mailbox time last read" when there is a quota error for a maildir
5835 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5837 35. Added +ignore_unknown as a special item in host lists.
5839 36. Code for decoding IPv6 addresses in host lists is now included, even if
5840 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5841 address was recognized as an IP address, but was then not correctly decoded
5842 into binary, causing unexpected and incorrect effects when compared with
5849 1. Very minor rewording of debugging text in manualroute to say "list of
5850 hosts" instead of "hostlist".
5852 2. If verify=header_syntax was set, and a header line with an unqualified
5853 address (no domain) and a large number of spaces between the end of the
5854 name and the colon was received, the reception process suffered a buffer
5855 overflow, and (when I tested it) crashed. This was caused by some obsolete
5856 code that should have been removed. The fix is to remove it!
5858 3. When running in the test harness, delay a bit after writing a bounce
5859 message to get a bit more predictability in the log output.
5861 4. Added a call to search_tidyup() just before forking a reception process. In
5862 theory, someone could use a lookup in the expansion of smtp_accept_max_
5863 per_host which, without the tidyup, could leave open a database connection.
5865 5. Added the variables $recipient_data and $sender_data which get set from a
5866 lookup success in an ACL "recipients" or "senders" condition, or a router
5867 "senders" option, similar to $domain_data and $local_part_data.
5869 6. Moved the writing of debug_print from before to after the "senders" test
5872 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5873 problems for message scanning, either using a data ACL, or using
5874 local_scan() because the Received: header was not generated till after they
5875 were called (in order to set the time as the time of reception completion).
5876 I have revised the way this works. The header is now generated after the
5877 body is received, but before the ACL or local_scan() are called. After they
5878 are run, the timestamp in the header is updated.
5884 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5885 before starting a queue runner without re-exec. This happened only when
5886 deliver_drop_privilege was set or when the Exim user was set to root. The
5887 effect of the bug was that timeouts during subsequent deliveries caused
5888 crashes instead of being properly handled. The handler is now left at its
5889 default (and expected) setting.
5891 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5892 message, again when deliver_drop_privilege is set or Exim is run as root.
5893 The bug described in (1) was not present in this case, but the tidying up
5894 of the other signals was missing. I have made the two cases consistent.
5896 3. The ignore_target_hosts setting on a manualroute router was being ignored
5897 for hosts that were looked up using the /MX notation.
5899 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5902 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5903 operated on the sender address. After changing the $sender_address to <>
5904 for the sender address verify, Exim was re-instated it as the original
5905 (before rewriting) address, but remembering that it had rewritten it, so it
5906 wasn't rewriting it again. This bug also had the effect of breaking the
5907 sender address verification caching when the sender address was rewritten.
5909 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5910 This has been changed so that if the ip literal address matches
5911 ignore_target_hosts, the router declines.
5913 7. Added expansion conditions match_domain, match_address, and match_local_
5914 part (NOT match_host).
5916 8. The placeholder for the Received: header didn't have a length field set.
5918 9. Added code to Exim itself and to exim_lock to test for a specific race
5919 condition that could lead to file corruption when using MBX delivery. The
5920 issue is with the lockfile that is created in /tmp. If this file is removed
5921 after a process has opened it but before that process has acquired a lock,
5922 there is the potential for a second process to recreate the file and also
5923 acquire a lock. This could lead to two Exim processes writing to the file
5924 at the same time. The added code performs the same test as UW imapd; it
5925 checks after acquiring the lock that its file descriptor still refers to
5926 the same named file.
5928 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5929 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5932 11. Added the smtp_active_hostname option. If used, this will typically be made
5933 to depend on the incoming interface address. Because $interface_address is
5934 not set up until the daemon has forked a reception process, error responses
5935 that can happen earlier (such as "too many connections") no longer contain
5938 12. If an expansion in a condition on a "warn" statement fails because a lookup
5939 defers, the "warn" statement is abandoned, and the next ACL statement is
5940 processed. Previously this caused the whole ACL to be aborted.
5942 13. Added the iplsearch lookup type.
5944 14. Added ident_timeout as a log selector.
5946 15. Added tls_certificate_verified as a log selector.
5948 16. Added a global option tls_require_ciphers (compare the smtp transport
5949 option of the same name). This controls incoming TLS connections.
5951 17. I finally figured out how to make tls_require_ciphers do a similar thing
5952 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5953 before starting the TLS session.
5955 18. Tabs are now shown as \t in -bP output.
5957 19. If the log selector return_path_on_delivery was set, Exim crashed when
5958 bouncing a message because it had too many Received: header lines.
5960 20. If two routers both had headers_remove settings, and the first one included
5961 a superfluous trailing colon, the final name in the first list and the
5962 first name in the second list were incorrectly joined into one item (with a
5963 colon in the middle).
5969 1. Added -C and -D options to the exinext utility, mainly to make it easier
5970 to include in the automated testing, but these could be helpful when
5971 multiple configurations are in use.
5973 2. The exinext utility was not formatting the output nicely when there was
5974 an alternate port involved in the retry record key, nor when there was a
5975 message id as well (for retries that were specific to a specific message
5976 and a specific host). It was also confused by IPv6 addresses, because of
5977 the additional colons they contain. I have fixed the IPv4 problem, and
5978 patched it up to do a reasonable job for IPv6.
5980 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5981 delivery, the log line now contains "pipelined" if PIPELINING was used.
5983 4. An SMTP transport process used to panic and die if the bind() call to set
5984 an explicit outgoing interface failed. This has been changed; it is now
5985 treated in the same way as a connect() failure.
5987 5. A reference to $sender_host_name in the part of a conditional expansion
5988 that was being skipped was still causing a DNS lookup. This no longer
5991 6. The def: expansion condition was not recognizing references to header lines
5992 that used bh_ and bheader_.
5994 7. Added the _cache feature to named lists.
5996 8. The code for checking quota_filecount in the appendfile transport was
5997 allowing one more file than it should have been.
5999 9. For compatibility with Sendmail, the command line option
6007 and sets the incoming protocol and host name (for trusted callers). The
6008 host name and its colon can be omitted when only the protocol is to be set.
6009 Note the Exim already has two private options, -pd and -ps, that refer to
6010 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6011 "s", but I don't think that's a major issue.
6013 10. A number of refactoring changes to the code, none of which should affect
6016 (a) The number of logging options was getting close to filling up the
6017 32-bit word that was used as a bit map. I have split them into two classes:
6018 those that are passed in the argument to log_write(), and those that are
6019 only ever tested independently outside of that function. These are now in
6020 separate 32-bit words, so there is plenty of room for expansion again.
6021 There is no change in the user interface or the logging behaviour.
6023 (b) When building, for example, log lines, the code previously used a
6024 macro that called string_cat() twice, in order to add two strings. This is
6025 not really sufficiently general. Furthermore, there was one instance where
6026 it was actually wrong because one of the argument was used twice, and in
6027 one call a function was used. (As it happened, calling the function twice
6028 did not affect the overall behaviour.) The macro has been replaced by a
6029 function that can join an arbitrary number of extra strings onto a growing
6032 (c) The code for expansion conditions now uses a table and a binary chop
6033 instead of a serial search (which was left over from when there were very
6034 few conditions). Also, it now recognizes conditions like "pam" even when
6035 the relevant support is not compiled in: a suitably worded error message is
6036 given if an attempt is made to use such a condition.
6038 11. Added ${time_interval:xxxxx}.
6040 12. A bug was causing one of the ddress fields not to be passed back correctly
6041 from remote delivery subprocesses. The field in question was not being
6042 subsequently used, so this caused to problems in practice.
6044 13. Added new log selectors queue_time and deliver_time.
6046 14. Might have fixed a bug in maildirsizefile handling that threw up
6047 "unexpected character" debug warnings, and recalculated the data
6048 unnecessarily. In any case, I expanded the warning message to give more
6051 15. Added the message "Restricted characters in address" to the statements in
6052 the default ACL that block characters like @ and % in local parts.
6054 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6055 Three changes have been made:
6057 (a) There was a serious bug; a negative response to MAIL caused the whole
6058 recipient domain to be cached as invalid, thereby blocking all messages
6059 to all local parts at the same domain, from all senders. This bug has
6060 been fixed. The domain is no longer cached after a negative response to
6061 MAIL if the sender used is not empty.
6063 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6066 (c) A new callout option, "use_sender" has been added for people who want
6067 the modified behaviour.
6073 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6076 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6077 indeed breaks things for older releases.
6079 3. Added additional logging to the case where there is a problem reading data
6080 from a filter that is running in a subprocess using a pipe, in order to
6081 try to track down a specific problem.
6083 4. Testing facility fudge: when running in the test harness and attempting
6084 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6085 getting "No route to host". Convert this to a timeout.
6087 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6090 6. Some OS don't have socklen_t but use size_t instead. This affects the
6091 fifth argument of getsockopt() amongst other things. This is now
6092 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6093 can be set for individual OS. I have set it for SunOS5, OSF1, and
6094 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6095 some earlier ones do not.
6097 7. Change 4.30/15 was not doing the test caselessly.
6099 8. The standard form for an IPv6 address literal was being rejected by address
6100 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6101 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6102 this, as well as the form without the "IPv6" on the front (but only when
6103 address literals are enabled, of course).
6105 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6107 10. Exim crashed if a message with an empty sender address specified by -f
6108 encountered a router with an errors_to setting. This could be provoked only
6109 by a command such as
6113 where an empty string was supplied; "<>" did not hit this bug.
6115 11. Installed PCRE release 4.5.
6117 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6118 remained set. It is now erased.
6120 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6121 times from message ids (which are base 36 rather than the normal 62).
6123 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6124 were being counted as actual protocol errors, and logged if the log
6125 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6126 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6127 and DATA following a set of rejected RCPTs do not count as protocol errors.
6128 In other words, Exim assumes they were pipelined, though this may not
6129 actually be the case. Of course, in all cases the client gets an
6130 appropriate error code.
6132 15. If a lookup fails in an ACL condition, a message about the failure may
6133 be available; it is used if testing the ACL cannot continue, because most
6134 such messages specify what the cause of the deferral is. However, some
6135 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6136 that caused an old message to be retained and used if a later statement
6137 caused a defer, replacing the real cause of the deferral.
6139 16. If an IP address had so many PTR records that the DNS lookup buffer
6140 was not large enough to hold them, Exim could crash while trying to process
6141 the truncated data. It now detects and logs this case.
6143 17. Further to 4.21/58, another change has been made: if (and only if) the
6144 first line of a message (the first header line) ends with CRLF, a bare LF
6145 in a subsequent header line has a space inserted after it, so as not to
6146 terminate the header.
6148 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6149 unnecessarily, used atoi() instead of strtol(), and didn't check the
6150 termination when getting file sizes from file names by regex.
6152 19. Completely re-implemented the support for maildirsize files, in the light
6153 of a number of problems with the previous contributed implementation
6154 (4.30/29). In particular:
6156 . If the quota is zero, the maildirsize file is maintained, but no quota is
6159 . If the maildir directory does not exist, it is created before any attempt
6160 to write a maildirsize file.
6162 . The quota value in the file is just a cache; if the quota is changed in
6163 the transport, the new value overrides.
6165 . A regular expression is available for excluding directories from the
6168 20. The autoreply transport checks the characters in options that define the
6169 message's headers; it allows continued headers, but it was checking with
6170 isspace() after an embedded newline instead of explicitly looking for a
6173 21. If all the "regular" hosts to which an address was routed had passed their
6174 expiry times, and had not reached their retry times, the address was
6175 bounced, even if fallback hosts were defined. Now Exim should go on to try
6178 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6179 equivalent code in the SMTP transport. Some hosts send humungous responses
6180 to HELO/EHLO, more than 1024 it seems.
6182 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6183 gives compiler warnings in some environments. I've now done it "properly",
6186 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6187 (because of problems with the built-in one) was declared to return uschar *
6188 instead of char *, causing compiler failure.
6190 25. Fixed a file descriptor leak when processing alias/forward files.
6192 26. Fixed a minor format string issue in dbfn.c.
6194 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6196 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6197 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6200 29. When a sender address is verified, it is cached, to save repeating the test
6201 when there is more than one recipient in a message. However, when the
6202 verification involves a callout, it is possible for different callout
6203 options to be set for different recipients. It is too complicated to keep
6204 track of this in the cache, so now Exim always runs a verification when a
6205 callout is required, relying on the callout cache for the optimization.
6206 The overhead is duplication of the address routing, but this should not be
6209 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6210 to be verified with callout=postmaster, and the main callout worked but the
6211 postmaster check failed, the verification correctly failed. However, if a
6212 subsequent RCPT command asked for sender verification *without* the
6213 postmaster check, incorrect caching caused this verification also to fail,
6216 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6217 it was not caching the DNS options (qualify_single, search_parents) that
6218 were used when the lookup failed. A subsequent lookup with different
6219 options therefore always gave the same answer, though there were cases
6220 where it should not have. (Example: a "domains = !$mx_any" option on a
6221 dnslookup router: the "domains" option is always processed without any
6222 widening, but the router might have qualify_single set.) Now Exim uses the
6223 cached value only when the same options are set.
6225 32. Added John Jetmore's "exipick" utility to the distribution.
6227 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6228 than a timeout (e.g. a certificate is required, and is not provided), an
6229 Exim server now closes the connection immediately. Previously it waited for
6230 the client to close - but if the client is SSL, it seems that they each
6231 wait for each other, leading to a delay before one of them times out.
6233 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6234 maintained 0.8.x compatibility because I don't think many are using it, and
6235 it is clearly obsolete.
6237 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6240 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6241 client certificate was expired. A simple patch fixes this, though I don't
6242 understand the full logic of why the verify callback is called multiple
6245 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6246 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6247 which causes problems with some clients (such as the Certicom SSL Plus
6248 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6249 disables the coutermeasure allowing Eudora to connect."
6251 38. Exim was not checking that a write() to a log file succeeded. This could
6252 lead to Bad Things if a log got too big, in particular if it hit a file
6253 size limit. Exim now panics and dies if it cannot write to a log file, just
6254 as it does if it cannot open a log file.
6256 39. Modified OS/Makefile-Linux so that it now contains
6258 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6260 The two -D definitions ensure that Exim is compiled with large file
6261 support, which makes it possible to handle log files that are bigger than
6264 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6265 instance) a domain was checked against a named list that involved a lookup,
6266 causing $domain_data to be set, then another domain was checked against the
6267 same list, then the first domain was re-checked, the value of $domain_data
6268 after the final check could be wrong. In particular, if the second check
6269 failed, it could be set empty. This bug probably also applied to
6272 41. The strip_trailing_dot option was not being applied to the address given
6273 with the -f command-line option.
6275 42. The code for reading a message's header from the spool was incrementing
6276 $received_count, but never initializing it. This meant that the value was
6277 incorrect (doubled) while delivering a message in the same process in which
6278 it was received. In the most common configuration of Exim, this never
6279 happens - a fresh exec is done - but it can happen when
6280 deliver_drop_privilege is set.
6282 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6283 it now includes up to 150 characters of the unexpected data in the log
6286 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6287 and building data strings. The size of both of these buffers was 10 000
6288 bytes - far larger than anybody would *ever* want, thought I. Needless to
6289 say, somebody hit the limit. I have increased the maximum line length to
6290 20 000 and the maximum data length of concatenated lines to 100 000. I have
6291 also fixed two bugs, because there was no checking on these buffers. Tsk,
6292 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6293 buffer is too small.
6295 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6296 lsearch lookups. Now it does.
6298 46. When parsing a route_list item in a manualroute router, a fixed-length
6299 buffer was used for the list of hosts. I made this 1024 bytes long,
6300 thinking that nobody would ever have a list of hosts that long. Wrong.
6301 Somebody had a whole pile of complicated expansion conditions, and the
6302 string was silently truncated, leading to an expansion error. It turns out
6303 that it is easier to change to an unlimited length (owing to other changes
6304 that have happened since this code was originally written) than to build
6305 structure for giving a limitation error. The length of the item that
6306 expands into the list of hosts is now unlimited.
6308 47. The lsearch lookup could not handle data where the length of text line was
6309 more than 4095 characters. Such lines were truncated, leading to shortened
6310 data being returned. It should now handle lines of any length.
6312 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6313 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6316 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6317 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6318 no longer happens in comment lines. A list of replacements is now placed
6319 at the head of all of the source files, except those whose only change is
6320 to replace PERL_COMMAND in the very first #! line.
6322 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6323 messages on the queue, with a bottom-up merge sort, using code contributed
6324 by Michael Haardt. This should make operations like -bp somewhat faster on
6325 large queues. It won't affect queue runners, except when queue_run_in_order
6328 51. Installed eximstats 1.31 in the distribution.
6330 52. Added support for SRV lookups to the dnslookup router.
6332 53. If an ACL referred to $message_body or $message_body_end, the value was not
6333 reset for any messages that followed in the same SMTP session.
6335 54. The store-handling optimization for building very long strings was not
6336 differentiating between the different store pools. I don't think this
6337 actually made any difference in practice, but I've tidied it.
6339 55. While running the routers to verify a sender address, $sender_address
6340 was still set to the sender address. This is wrong, because when routing to
6341 send a bounce to the sender, it would be empty. Therefore, I have changed
6342 it so that, while verifying a sender address, $sender_address is set to <>.
6343 (There is no change to what happens when verifying a recipient address.)
6345 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6346 target A or AAAA records (if not already returned) without resetting the
6347 qualify_single or search_parents options of the DNS resolver. These are
6348 inappropriate in this case because the targets of MX and SRV records must
6349 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6350 target that, when qualified, matched something in the local domain. These
6351 two options are now turned off when doing these lookups.
6353 57. It seems that at least some releases of Reiserfs (which does not have the
6354 concept of a fixed number of inodes) returns zero and not -1 for the
6355 number of available inodes. This interacted badly with check_spool_inodes,
6356 which assumed that -1 was the "no such thing" setting. What I have done is
6357 to check that the total number of inodes is greater than zero before doing
6358 the test of how many are available.
6360 58. When a "warn" ACL statement has a log_message modifier, the message is
6361 remembered, and not repeated. This is to avoid a lot of repetition when a
6362 message has many recipients that cause the same warning to be written.
6363 However, Exim was preserving the list of already written lines for an
6364 entire SMTP session, which doesn't seem right. The memory is now reset if a
6365 new message is started.
6367 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6368 debugging output unless log_rewrite was also set.
6370 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6371 of (int)(handle) when we know that handle contains (void *)(-1).
6373 61. The Exim daemon panic-logs an error return when it closes the incoming
6374 connection. However "connection reset by peer" seems to be common, and
6375 isn't really an error worthy of noting specially, so that particular error
6378 62. When Exim is trying to find all the local interfaces, it used to panic and
6379 die if the ioctl to get the interface flags failed. However, it seems that
6380 on at least one OS (Solaris 9) it is possible to have an interface that is
6381 included in the list of interfaces, but for which you get a failure error
6382 for this call. This happens when the interface is not "plumbed" into a
6383 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6384 failure of the "get flags" call assumes that the interface is down.
6386 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6387 makes life easier for people who are doing arithmetic on fields extracted
6388 from dates, where you often get leading zeros that should not be
6389 interpreted as octal.
6391 64. Added qualify_domain to the redirect router, to override the global
6394 65. If a pathologically long header line contained very many addresses (the
6395 report of this problem mentioned 10 000) and each of them was rewritten,
6396 Exim could use up a very large amount of memory. (It kept on making new
6397 copies of the header line as it rewrote, and never released the old ones.)
6398 At the expense of a bit more processing, the header rewriting function has
6399 been changed so that it no longer eats memory in this way.
6401 66. The generation of the Received: header has been moved from the time that a
6402 message starts to be received, to the time that it finishes. The timestamp
6403 in the Received: header should now be very close to that of the <= log
6404 line. There are two side-effects of this change:
6406 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6407 logged header lines no longer include the local Received: line, because
6408 it has not yet been created. The same applies to a copy of the message
6409 that is returned to a non-SMTP sender when a message is rejected.
6411 (b) When a filter file is tested using -bf, no additional Received: header
6412 is added to the test message. After some thought, I decided that this
6415 This change does not affect the value of $received_for. It is still set
6416 after address rewriting, but before local_scan() is called.
6418 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6420 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6421 gave an unhelpful panic error message, and a defer error. I have managed to
6422 change this behaviour so that it now rejects any supplied certificate,
6423 which seems right, as the list of acceptable certificates is empty.
6425 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6426 gave an unhelpful defer error. I have not managed to make this reject any
6427 supplied certificates, but the error message it gives is "no certificate
6428 supplied", which is not helpful.
6430 70. exigrep's output now also includes lines that are not associated with any
6431 message, but which match the given pattern. Implemented by a patch from
6432 Martin Sluka, which also tidied up the Perl a bit.
6434 71. Recipient callout verification, like sender verification, was using <> in
6435 the MAIL FROM command. This isn't really the right thing, since the actual
6436 sender may affect whether the remote host accepts the recipient or not. I
6437 have changed it to use the actual sender in the callout; this means that
6438 the cache record is now keyed on a recipient/sender pair, not just the
6439 recipient address. There doesn't seem to be a real danger of callout loops,
6440 since a callout by the remote host to check the sender would use <>.
6441 [SEE ABOVE: changed after hitting problems.]
6443 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6444 temporary errors. However, in the case of such a code being given after
6445 the end of a data transmission (i.e. after ".") Exim was failing to write
6446 a retry record for the message. (Yes, there was some broken host that was
6447 actually sending 8xx at this point.)
6449 73. An unknown lookup type in a host list could cause Exim to panic-die when
6450 the list was checked. (An example that provoked this was putting <; in the
6451 middle of a list instead of at the start.) If this happened during a DATA
6452 ACL check, a -D file could be left lying around. This kind of configuration
6453 error no longer causes Exim to die; instead it causes a defer error. The
6454 incident is still logged to the main and panic logs.
6456 74. Buglet left over from Exim 3 conversion. The message "too many messages
6457 in one connection" was written to the rejectlog but not the mainlog, except
6458 when address rewriting (yes!) was being logged.
6460 75. Added write_rejectlog option.
6462 76. When a system filter was run not as root (that is, when system_filter_user
6463 was set), the values of the $n variables were not being returned to the
6464 main process; thus, they were not subsequently available in the $sn
6467 77. Added +return_path_on_delivery log selector.
6469 78. A connection timeout was being treated differently from recipients deferred
6470 when testing hosts_max_try with a message that was older than the host's
6471 retry timeout. (The host should not be counted, thus allowing all hosts to
6472 be tried at least once before bouncing.) This may have been the cause of an
6473 occasionally reported bug whereby a message would remain on the queue
6474 longer than the retry timeout, but would be bounced if a delivery was
6475 forced. I say "may" because I never totally pinned down the problem;
6476 setting up timeout/retry tests is difficult. See also the next item.
6478 79. The ultimate address timeout was not being applied to errors that involved
6479 a combination of host plus message (for example, a timeout on a MAIL
6480 command). When an address resolved to a number of possible hosts, and they
6481 were not all tried for each delivery (e.g. because of hosts_max_try), a
6482 message could remain on the queue longer than the retry timeout.
6484 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6487 81. Fixed an obscure SMTP outgoing bug which required at least the following
6488 conditions: (a) there was another message waiting for the same server;
6489 (b) the server returned 5xx to all RCPT commands in the first message so
6490 that the message was not completed; (c) the server dropped the connection
6491 or gave a negative response to the RSET that Exim sends to abort the
6492 transaction. The observed case was a dropped connection after DATA that had
6493 been sent in pipelining mode. That is, the server had advertised PIPELINING
6494 but was not implementing it correctly. The effect of the bug was incorrect
6495 behaviour, such as trying another host, and this could lead to a crash.
6501 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6502 and daemon.c were passed as pointers to ints; they should have been
6503 pointers to socklen_t variables (which are typically unsigned ints).
6505 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6508 3. Fixed a really odd bug that affected only the testing scheme; patching a
6509 certain fixed string in the binary changed the value of another string that
6510 happened to be identical to the end of the original first string.
6512 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6513 name", it returns that address as the IP address. On some operating
6514 systems (e.g. Solaris), it also passes back the IP address string as the
6515 "host name". However, on others (e.g. Linux), it passes back an empty
6516 string. Exim wasn't checking for this, and was changing the host name to an
6517 empty string, assuming it had been canonicalized.
6519 5. Although rare, it is permitted to have more than one PTR record for a given
6520 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6521 all the names associated with an address, because they do in Solaris.
6522 However, it seems that they do not in Linux for data that comes from the
6523 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6524 I found this out when I moved to a new Linux workstation and tried to run
6525 the Exim test suite.
6527 To get round this problem I have changed the code so that it now does its
6528 own call to the DNS to look up PTR records when searching for a host name.
6529 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6530 addresses that are only in /etc/hosts are still found.
6532 This behaviour is, however, controlled by an option called host_lookup_
6533 order, which defaults to "bydns:byaddr". If people want to use the other
6534 order, or indeed, just use one or the other means of lookup, they can
6535 specify it in this variable.
6537 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6538 some operating systems, this comes back from gethostbyaddr() as an empty
6539 string, and this is what Exim used to test for. However, it seems that in
6540 other systems, "." is yielded. Exim now tests for this case too.
6542 7. The values of check_spool_space and check_log_space are now held internally
6543 as a number of kilobytes instead of an absolute number of bytes. If a
6544 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6545 kilobyte. This means that much larger values can be stored.
6547 8. Exim monitor: an attempt to get the action menu when not actually pointing
6548 at a message produces an empty menu entitled "No message selected". This
6549 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6550 no entries in it ("Shell widget menu has zero width and/or height"). So I
6551 have added a single, blank menu entry in this case.
6553 9. Added ${quote_local_part.
6555 10. MIME decoding is now applied to the contents of Subject: header lines when
6558 11. Now that a reference to $sender_host_address automatically causes a reverse
6559 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6560 host lookup before query-style lookups in lists that might use this
6561 variable. This has therefore been abolished, and the "net-" prefix is no
6562 longer necessary for query-style lookups.
6564 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6565 have been a typo for LFLAGS, so it has been changed.
6567 13. The install script calls Exim with "-C /dev/null" in order to find the
6568 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6569 to be output. However, since Exim outputs its version number before the
6570 error, it didn't break the script. It just looked ugly. I fixed this by
6571 always allowing "-C /dev/null" if the caller is root.
6573 14. Ignore overlarge ACL variable number when reading spool file - insurance
6574 against a later release with more variables having written the file.
6576 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6577 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6578 this, as well as the form without the "IPv6" on the front.
6580 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6581 OS/Makefile-Darwin file.
6583 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6584 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6586 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6587 variable. However, some OS use ints for inodes. I've added cast to long int
6588 to get rid of the compiler warning.
6590 19. I had forgotten to lock out "/../" in configuration file names when
6591 ALT_CONFIG_PREFIX was set.
6593 20. Routers used for verification do not need to specify transports. However,
6594 if such a router generated a host list, and callout was configured, Exim
6595 crashed, because it could not find a port number from the (non-existent)
6596 transport. It now assumes port 25 in this circumstance.
6598 21. Added the -t option to exigrep.
6600 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6601 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6602 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6603 them is set, LOOKUP_LSEARCH is forced.
6605 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6607 24. Added sender and host information to the "rejected by local_scan()" log
6608 line; previously there was no indication of these.
6610 25. Added .include_if_exists.
6612 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6613 Linux. It turns out that not all file systems support this. Apparently some
6614 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6615 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6616 not supported on the file descriptor, is now ignored when Exim is trying to
6617 sync a directory. This applies only to Linux.
6619 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6621 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6622 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6625 29. Added support for maildirsize files from supplied patch (modified a bit).
6627 30. The use of :fail: followed by an empty string could lead Exim to respond to
6628 sender verification failures with (e.g.):
6630 550 Verification failed for <xxx>
6631 550 Sender verify failed
6633 where the first response line was missing the '-' that indicates it is not
6634 the final line of the response.
6636 31. The loop for finding the name of the user that called Exim had a hardwired
6637 limit of 10; it now uses the value of finduser_retries, which is used for
6638 all other user lookups.
6640 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6643 33. Exim was neglecting to zero errno before one call of strtol() when
6644 expanding a string and expecting an integer value. On some systems this
6645 resulted in spurious "integer overflow" errors. Also, it was casting the
6646 result into an int without checking.
6648 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6649 did not work. The code looks as if it has *never* worked, though it appears
6650 to have been documented since at least release 1.62. I have made it work.
6652 35. The "timeout_DNS" error in retry rules, also documented since at least
6653 1.62, also never worked. As it isn't clear exactly what this means, and
6654 clearly it isn't a major issue, I have abolished the feature by treating it
6655 as "timeout", and writing a warning to the main and panic logs.
6657 36. The display of retry rules for -brt wasn't always showing the error code
6660 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6661 timeout_connect_A, timeout_connect_MX.
6663 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6664 to the empty sender.
6666 39. The daemon was not analysing the content of -oX till after it had closed
6667 stderr and disconnected from the controlling terminal. This meant that any
6668 syntax errors were only noted on the panic log, and the return code from
6669 the command was 0. By re-arranging the code a little, I've made the
6670 decoding happen first, so such errors now appear on stderr, and the return
6671 code is 1. However, the actual setting up of the sockets still happens in
6672 the disconnected process, so errors there are still only recorded on the
6675 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6676 connections (as happens on some IP stacks) was logged at start up time as
6677 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6678 it from "IPv6 and IPv4", which means that two separate sockets are being
6681 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6682 says whether AF_INET or AF_INET6 was passed as an argument.
6684 42. Exiwhat output was messed up when time zones were included in log
6687 43. Exiwhat now gives more information about the daemon's listening ports,
6688 and whether -tls-on-connect was used.
6690 44. The "port" option of the smtp transport is now expanded.
6692 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6693 silently ignored. Now an error message is written to the main and panic
6696 46. There's a new ACL modifier called "logwrite" which writes to a log file
6697 as soon as it is encountered.
6699 47. Added $local_user_uid and $local_user_gid at routing time.
6701 48. Exim crashed when trying to verify a sender address that was being
6704 49. Exim was recognizing only a space character after ".include". It now also
6705 recognizes a tab character.
6707 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6708 extracting the relevant information from the specification. The man page no
6709 longer contains scrambled data for the -d option, and I've added a section
6710 at the front about calling Exim under different names.
6712 51. Added "extra_headers" argument to the "mail" command in filter files.
6714 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6717 53. Installed eximstats 1.29.
6719 54. Added transport_filter_timeout as a generic transport option.
6721 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6722 Cc: header lines. This was required by RFC 822, but it not required by RFC
6725 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6726 incoming messages that did not have them. Now it does so only if the
6727 message originates locally, that is, if there is no associated remote host
6728 address. When Resent- header lines are present, this applies to the Resent-
6729 lines rather than the non-Resent- lines.
6731 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6732 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6734 58. Messages for configuration errors now include the name of the main
6735 configuration file - useful now that there may be more than one file in a
6736 list (.included file names were always shown).
6738 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6739 for those rare installations that do not start the daemon as root or run it
6740 setuid root. I've cut out the call to initgroups() if the daemon is not
6743 60. The Exim user and group can now be bound into the binary as text strings
6744 that are looked up at the start of Exim's processing.
6746 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6748 62. Added $mailstore_basename variable.
6750 63. Installed patch to sieve.c from Michael Haardt.
6752 64. When Exim failed to open the panic log after failing to open the main log,
6753 the original message it was trying to log was written to stderr and debug
6754 output, but if they were not available (the usual case in production), it
6755 was lost. Now it is written to syslog before the two lines that record the
6756 failures to open the logs.
6758 65. Users' Exim filters run in subprocesses under the user's uid. It is
6759 possible for a "deliver" command or an alias in a "personal" command to
6760 provoke an address rewrite. If logging of address rewriting is configured,
6761 this fails because the process is not running as root or exim. There may be
6762 a better way of dealing with this, but for the moment (because 4.30 needs
6763 to be released), I have disabled address rewrite logging when running a
6764 filter in a non-root, non-exim process.
6770 1. The buildconfig auxiliary program wasn't quoting the value set for
6771 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6772 not defined. This bug was masked in 4.22 by the effect that was fixed in
6775 2. Some messages that were rejected after a message id was allocated were
6776 shown as "incomplete" by exigrep. It no longer does this for messages that
6777 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6779 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6780 have allow_domain_literals set, the ID did not get logged in the <= line.
6781 Domain literals are now always recognized in Message-ID: header lines.
6783 4. The first argument for a ${extract expansion item is the key name or field
6784 number. Leading and trailing spaces in this item were not being ignored,
6785 causing some misleading effects.
6787 5. When deliver_drop_privilege was set, single queue runner processes started
6788 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6789 same command in the process it spins off) were not dropping privilege.
6791 6. When the daemon running as "exim" started a queue runner, it always
6792 re-executed Exim in the spun-off process. This is a waste of effort when
6793 deliver_drop_privilege is set. The new process now just calls the
6794 queue-runner function directly.
6800 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6803 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6804 ignored. Though the use of -oP was forcing the writing of a pid file, it
6805 was always written to the default place.
6807 3. If the message "no IP address found for host xxxx" is generated during
6808 incoming verification, it is now followed by identification of the incoming
6809 connection (so you can more easily find what provoked it).
6811 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6813 5. Added some features to "harden" Exim a bit more against certain attacks:
6815 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6816 be put in Local/Makefile. This is like the never_users runtime option,
6817 but it cannot be overridden. The default setting is "root".
6819 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6820 prefix string with which any file named in a -C command line option
6823 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6824 is retained for -C and -D only if the caller of Exim is root. Without
6825 it, the exim user may also use -C and -D and retain privilege.
6827 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6828 command line option is disabled.
6830 6. Macro names set by the -D option must start with an upper case letter, just
6831 like macro names defined in the configuration file.
6833 7. Added "dereference=" facility to LDAP.
6835 8. Two instances of the typo "uknown" in the source files are fixed.
6837 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6838 the Configure-Makefile script screwed up while processing it.
6840 10. Incorporated PCRE 4.4.
6842 11. The SMTP synchronization check was not operating right at the start of an
6843 SMTP session. For example, it could not catch a HELO sent before the client
6844 waited for the greeting. There is now a check for outstanding input at the
6845 point when the greeting is written. Because of the duplex, asynchronous
6846 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6847 way, but not yet received, when the check is performed.
6849 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6850 on TCP/IP sockets, because this apparently causes some broken clients to
6853 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6854 unchanged) from the Cygwin maintainer.
6856 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6857 was supported instead of testing for mbx. Effectively a typo.
6859 15. The spa authenticator server code was not checking that the input it
6860 received was valid base64.
6862 16. The debug output line for the "set" modifier in ACLs was not showing the
6863 name of the variable that was being set.
6865 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6867 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6868 lookup on its own. Something else was needed to trigger the lookup. For
6869 example, a match in host_lookup or the need for a host name in a host list.
6870 Now, if $sender_host_name is referenced and the host name has not yet been
6871 looked up, a lookup is performed. If the lookup fails, the variable remains
6872 empty, and $host_lookup_failed is set to "1".
6874 19. Added "eqi" as a case-independent comparison operator.
6876 20. The saslauthd authentication condition could segfault if neither service
6877 nor realm was specified.
6879 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6880 error message that was logged was misleading, and incoming SMTP
6881 connections were dropped. The message is now more accurate, and temporary
6882 errors are given to SMTP connections.
6884 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6885 (except RSET) with a 421 error, until QUIT is received. However, it was
6886 failing to send a response to QUIT.
6888 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6889 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6890 ACL is now run after the helo_try_verify_hosts code.
6892 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6893 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6894 case-independent, so other case variants are also recognized). Apparently
6895 some systems use these upper case variants.
6897 25. If more than two messages were waiting for the same host, and a transport
6898 filter was specified for the transport, Exim sent two messages over the
6899 same TCP/IP connection, and then failed with "socket operation on non-
6900 socket" when it tried to send the third.
6902 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6904 27. The extern definition of crypt16() in expand.c was not being excluded when
6905 the OS had its own crypt16() function.
6907 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6908 as a preferred synonym for return_size_limit, both as an option and as an
6911 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6913 30. Changed the default configuration ACL to relax the local part checking rule
6914 for addresses that are not in any local domains. For these addresses,
6915 slashes and pipe symbols are allowed within local parts, but the sequence
6916 /../ is explicitly forbidden.
6918 31. SPA server authentication was not clearing the challenge buffer before
6921 32. log_message in a "warn" ACL statement was writing to the reject log as
6922 well as to the main log, which contradicts the documentation and doesn't
6923 seem right (because no rejection is happening). So I have stopped it.
6925 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6926 However, I am unable to do any testing of this.
6928 34. Fixed an infelicity in the appendfile transport. When checking directories
6929 for a mailbox, to see if any needed to be created, it was accidentally
6930 using path names with one or more superfluous leading slashes; tracing
6931 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6933 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6934 contents are added to the log line that is written for every discarded
6935 recipient. (Previously a log_message setting was ignored.)
6937 36. The ${quote: operator now quotes the string if it is empty.
6939 37. The install script runs exim in order to find its version number. If for
6940 some reason other than non-existence or emptiness, which it checks, it
6941 could not run './exim', it was installing it with an empty version number,
6942 i.e. as "exim-". This error state is now caught, and the installation is
6945 38. An argument was missing from the function that creates an error message
6946 when Exim fails to connect to the socket for saslauthd authentication.
6947 This could cause Exim to crash, or give a corrupted message.
6949 39. Added isip, isip4, and isip6 to ${if conditions.
6951 40. The ACL variables $acl_xx are now saved with the message, and can be
6952 accessed later in routers, transports, and filters.
6954 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6955 strings in the file are not string-expanded.
6957 42. If a MAIL command specified a SIZE value that was too large to fit into an
6958 int variable, the check against message_size_limit failed. Such values are
6959 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6960 day this will have to be increased, but I don't think I want to be around
6961 when emails are that large.
6968 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6969 iconv() is not standard in FreeBSD.
6971 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6972 IPv6 enabled. The observed symptom was a segmentation fault on return from
6973 the function os_common_find_running_interfaces() in src/os.c.
6975 3. In the check_special_case() function in daemon.c I had used "errno" as an
6976 argument name, which causes warnings on some systems. This was basically a
6977 typo, since it was named "eno" in the comments!
6979 4. The code that waits for the clock to tick (at a resolution of some fraction
6980 of a second) so as to ensure message-id uniqueness was always waiting for
6981 at least one whole tick, when it could have waited for less. [This is
6982 almost certainly not relevant at current processor speeds, where it is
6983 unlikely to ever wait at all. But we try to future-proof.]
6985 5. The function that sleeps for a time interval that includes fractions of a
6986 second contained a race. It did not block SIGALRM between setting the
6987 timer, and suspending (a couple of lines later). If the interval was short
6988 and the sigsuspend() was delayed until after it had expired, the suspension
6989 never ended. On busy systems this could lead to processes getting stuck for
6992 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6993 process, before it forks any delivery processes. The open lookup caching
6994 mechanism meant that the open file or database connection was passed into
6995 the delivery process. The problem was that delivery processes always tidy
6996 up cached lookup data. This could cause a problem for the next delivery
6997 process started by the queue runner, because the external queue runner
6998 process does not know about the closure. So the next delivery process
6999 still has data in the lookup cache. In the case of a file lookup, there was
7000 no problem because closing a file descriptor in a subprocess doesn't affect
7001 the parent. However, if the lookup was caching a connection to a database,
7002 the connection was closed, and the second delivery process was likely to
7003 see errors such as "PGSQL: query failed: server closed the connection
7004 unexpectedly". The problem has been fixed by closing all cached lookups
7005 in a queue runner before running a delivery process.
7007 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7008 seem to have the "const" qualifier which it has on other OS. I've
7011 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7012 *of the same type* (client or server) with the same public name that an
7013 error should be diagnosed.
7015 9. When Exim looked up a host name for an IP address, but failed to find the
7016 original IP address when looking up the host name (a safety check), it
7017 output the message "<ip address> does not match any IP for NULL", which was
7018 confusing, to say the least. The bug was that the host name should have
7019 appeared instead of "NULL".
7021 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7022 user that is built into the binary, and the -C or -D options is used, root
7023 privilege is dropped before the configuration file is read. In addition,
7024 logging is switched to stderr instead of the normal log files. If the
7025 configuration then re-defines the Exim user, the unprivileged environment
7026 is probably not what is expected, so Exim logs a panic warning message (but
7029 However, if deliver_drop_privilege is set, the unprivileged state may well
7030 be exactly what is intended, so the warning has been cut out in that case,
7031 and Exim is allowed to try to write to its normal log files.
7037 1. smtp_return_error_details was not giving details for temporary sender
7038 or receiver verification errors.
7040 2. Diagnose a configuration error if two authenticators have the same public
7043 3. Exim used not to create the message log file for a message until the first
7044 delivery attempt. This could be confusing when incoming messages were held
7045 for policy or load reasons. The message log file is now created at the time
7046 the message is received, and an initial "Received" line is written to it.
7048 4. The automatically generated man page for command line options had a minor
7049 bug that caused no ill effects; however, a more serious problem was that
7050 the procedure for building the man page automatically didn't always
7051 operate. Consequently, release 4.20 contains an out-of-date version. This
7052 shouldn't happen again.
7054 5. When building Exim with embedded Perl support, the script that builds the
7055 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7056 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7058 6. The freeze_tell option was not being used for messages that were frozen on
7059 arrival, either by an ACL or by local_scan().
7061 7. Added the smtp_incomplete_transaction log selector.
7063 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7064 was accepting AUTH without a new EHLO.
7066 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7067 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7070 10. Logging of TCP/IP connections (when configured) now happens in the main
7071 daemon process instead of the child process, so that the TCP/IP connection
7072 count is more accurate (but it can never be perfect).
7074 11. The use of "drop" in a nested ACL was not being handled correctly in the
7075 outer ACL. Now, if condition failure induced by the nested "drop" causes
7076 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7077 or "require"), the connection is dropped.
7079 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7080 that yield "discard" can now be used with an "accept" or a "discard" verb,
7081 but an error is generated for any others (because I can't see a useful way
7082 to define what should happen).
7084 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7085 are now processed in the same way as lines in the Exim configuration file.
7086 In particular, continuation lines are supported.
7088 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7090 15. Added -ti meaning -t -i.
7092 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7093 domains, and warn by logging if others are found.
7095 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7096 returned from GIFCONF when Exim is trying to find the list of interfaces on
7097 a host. The code in os.c has been modified to copy each ifreq to an aligned
7098 structure in all cases.
7100 Also, in some cases, the returned ifreq's were being copied to a 'struct
7101 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7102 means the last couple of bytes of an IPv6 address could be chopped if the
7103 ifreq contained only a normal sockaddr (14 bytes storage).
7105 18. Named domain lists were not supported in the hosts_treat_as_local option.
7106 An entry such as +xxxx was not recognized, and was treated as a literal
7109 19. Ensure that header lines added by a DATA ACL are included in the reject log
7110 if the ACL subsequently rejects the message.
7112 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7113 MD5 (which is deprecated).
7115 21. When testing a filter file using -bf, Exim was writing a message when it
7116 took the sender from a "From " line in the message, but it was not doing so
7117 when it took $return_path from a Return-Path: header line. It now does.
7119 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7120 with a valid header line field name (a series of printing characters
7121 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7123 23. Changed "disc" in the source to "disk" to conform to the documentation and
7124 the book and for uniformity.
7126 24. Ignore Sendmail's -Ooption=value command line item.
7128 25. When execve() failed while trying to run a command in a pipe transport,
7129 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7130 could be confused with a return value of 69 from the command itself. This
7131 has been changed to 127, the value the shell returns if it is asked to run
7132 a non-existent command. The wording for the related log line suggests a
7133 non-existent command as the problem.
7135 26. If received_header_text expands to an empty string, do not add a Received:
7136 header line to the message. (Well, it adds a token one on the spool, but
7137 marks it "old" so that it doesn't get used or transmitted.)
7139 27. Installed eximstats 1.28 (addition of -nt option).
7141 28. There was no check for failure on the call to getsockname() in the daemon
7142 code. This can fail if there is a shortage of resources on the system, with
7143 ENOMEM, for example. A temporary error is now given on failure.
7145 29. Contrary to the C standard, it seems that in some environments, the
7146 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7147 program. Exim now does this explicitly; it affects the formatting of
7148 timestamps using strftime().
7150 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7151 complaints. I've now initialized all the variables, to avoid this.
7153 32. Header lines added by a system filter were not being "seen" during
7154 transport-time rewrites.
7156 33. The info_callback() function passed to OpenSSL is set up with type void
7157 (*)(SSL *, int, int), as described somewhere. However, when calling the
7158 function (actually a macro) that sets it up, the type void(*)() is
7159 expected. I've put in a cast to prevent warnings from picky compilers.
7161 34. If a DNS black list lookup found a CNAME record, but there were no A
7162 records associated with the domain it pointed at, Exim crashed.
7164 35. If a DNS black list lookup returned more than one A record, Exim ignored
7165 all but the first. It now scans all returned addresses if a particular IP
7166 value is being sought. In this situation, the contents of the
7167 $dnslist_value variable are a list of all the addresses, separated by a
7170 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7171 to accept a host name and all its aliases if the forward lookup for any of
7172 them yielded the IP address of the incoming connection. Now it accepts only
7173 those names whose forward lookup yields the correct IP address. Any other
7174 names are discarded. This closes a loophole whereby a rogue DNS
7175 administrator could create reverse DNS records to break through a
7176 wildcarded host restriction in an ACL.
7178 37. If a user filter or a system filter that ran in a subprocess used any of
7179 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7180 the wrong values were passed to the pipe command ($thisaddress had the
7181 value of $0, $0 had the value of $1, etc). This bug was introduced by
7182 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7184 38. Improved the line breaking for long SMTP error messages from ACLs.
7185 Previously, if there was no break point between 40 and 75 characters, Exim
7186 left the rest of the message alone. Two changes have been made: (a) I've
7187 reduced the minimum length to 35 characters; (b) if it can't find a break
7188 point between 35 and 75 characters, it looks ahead and uses the first one
7189 that it finds. This may give the occasional overlong line, but at least the
7190 remaining text gets split now.
7192 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7193 file descriptors might be low, and that setting 1000 would always raise it.
7194 It turns out that in some environments, the limit is already over 1000 and
7195 that lowering it causes trouble. So now Exim takes care not to decrease it.
7197 40. When delivering a message, the value of $return_path is set to $sender_
7198 address at the start of routing (routers may change the value). By an
7199 oversight, this default was not being set up when an address was tested by
7200 -bt or -bv, which affected the outcome if any router or filter referred to
7203 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7204 log only when "message" or "log_message" are set. However, if one of the
7205 conditions was an address verification, or a call to a nested ACL, the
7206 messages generated by the underlying test were being passed through. This
7207 no longer happens. The underlying message is available in $acl_verify_
7208 message for both "message" and "log_message" expansions, so it can be
7209 passed through if needed.
7211 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7212 new expansion $bh_ to give the encoded byte string without charset
7213 translation. Translation happens only if iconv() is available; HAVE_ICONV
7214 indicates this at build time. HEADERS_CHARSET gives the charset to
7215 translate to; headers_charset can change it in the configuration, and
7216 "headers charset" can change it in an individual filter file.
7218 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7219 that creates RFC 2047 encoded "words" labels them as that charset instead
7220 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7221 expansion operator; (ii) when Exim creates a From: line for a local
7222 message; (iii) when a header line is rewritten to include a "phrase" part.
7224 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7225 buggy, causing it to skip the first lines of messages whose message ID
7226 ended in 'D'. This would not have bitten before Exim release 4.14, because
7227 message IDs were unlikely to end in 'D' before then. The effect was to have
7228 incorrect size information for certain domains.
7230 45. #include "config.h" was missing at the start of the crypt16.c module. This
7231 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7234 46. If there was a timeout during a "random" callout check, Exim treated it as
7235 a failure of the random address, and carried on sending RSET and the real
7236 address. If the delay was just some slowness somewhere, the response to the
7237 original RCPT would be taken as a response to RSET and so on, causing
7238 mayhem of various kinds.
7240 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7241 when I implemented it. It didn't allow for the fact that some option values
7242 may legitimately be negative (e.g. size_addition), and it didn't even do
7243 the right test for positive values.
7245 48. Domain names in DNS records are case-independent. Exim always looks them up
7246 in lower case. Some resolvers return domain names in exactly the case they
7247 appear in the zone file, that is, they may contain uppercase letters. Not
7248 all resolvers do this - some return always lower case. Exim was treating a
7249 change of case by a resolver as a change of domain, similar to a widening
7250 of a domain abbreviation. This triggered its re-routing code and so it was
7251 trying to route what was effectively the same domain again. This normally
7252 caused routing to fail (because the router wouldn't handle the domain
7253 twice). Now Exim checks for this case specially, and just changes the
7254 casing of the domain that it ultimately uses when it transmits the message
7257 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7260 50. If a filter generated a file delivery with a non-absolute name (possible if
7261 no home directory exists for the router), the forbid_file option was not
7264 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7265 the existing equality matching.
7267 52. Exim was using ints instead of ino_t variables in some places where it was
7268 dealing with inode numbers.
7270 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7271 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7272 called TMPDIR, and if it finds it is different, it changes its value.
7274 54. The smtp_printf() function is now made available to local_scan() so
7275 additional output lines can be written before returning. There is also an
7276 smtp_fflush() function to enable the detection of a dropped connection.
7277 The variables smtp_input and smtp_batched_input are exported to
7280 55. Changed the default runtime configuration: the message "Unknown user"
7281 has been removed from the ACL, and instead placed on the localuser router,
7282 using the cannot_route_message feature. This means that any verification
7283 failures that generate their own messages won't get overridden. Similarly,
7284 the "Unrouteable address" message that was in the ACL for unverifiable
7285 relay addresses has also been removed.
7287 56. Added hosts_avoid_esmtp to the smtp transport.
7289 57. The exicyclog script was not checking for the esoteric option
7290 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7291 will work only if exicyclog is run under the appropriate euid.
7293 58. Following a discussion on the list, the rules by which Exim recognises line
7294 endings on incoming messages have been changed. The -dropcr and drop_cr
7295 options are now no-ops, retained only for backwards compatibility. The
7296 following line terminators are recognized: LF CRLF CR. However, special
7297 processing applies to CR:
7299 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7300 nor a local message in the state where . is a terminator.
7302 (ii) If a bare CR is encountered in a header line, an extra space is added
7303 after the line terminator so as not to end the header. The reasoning
7304 behind this is that bare CRs in header lines are most likely either
7305 to be mistakes, or people trying to play silly games.
7307 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7308 was being incorrectly given as 18 bytes larger than it should have been.
7309 This is a VOB (very old bug).
7311 60. This may never have affected anything current, but just in case it has:
7312 When the local host is found other than at the start of a list of hosts,
7313 the local host, those with the same MX, and any that follow, are discarded.
7314 When the list in question was part of a longer list of hosts, the following
7315 hosts (not currently being processed) were also being discarded. This no
7316 longer happens. I'm not sure if this situation could ever has previously
7319 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7322 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7323 header. This is something that is recommended in a new Internet Draft, and
7324 is something that is documented as being done by Sendmail. There are two
7325 possible values. For messages generated by the autoreply transport, Exim
7328 Auto-Submitted: auto-replied
7330 whereas for all other generated messages (e.g. bounces) it adds
7332 Auto-Submitted: auto-generated
7334 63. The "personal" condition in filters now includes a test for the
7335 Auto-Submitted: header. If it contains the string "auto-" the message it
7336 not considered personal.
7338 64. Added rcpt_include_affixes as a generic transport option.
7340 65. Added queue_only_override (default true).
7342 66. Added the syslog_duplication option.
7344 67. If what should have been the first header line of a message consisted of
7345 a space followed by a colon, Exim was mis-interpreting it as a header line.
7346 It isn't of course - it is syntactically invalid and should therefore be
7347 treated as the start of the message body. The misbehaviour could have
7348 caused a number of strange effects, including loss of data in subsequent
7349 header lines, and spool format errors.
7351 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7352 client host had authenticated. This control can now be exercised by an ACL
7353 for more flexibility.
7355 69. By default, callouts do not happen when testing with -bh. There is now a
7356 variant, -bhc, which does actually run the callout code, including
7357 consulting and updating the callout cache.
7359 70. Added support for saslauthd authentication, courtesy of Alexander
7362 71. If statvfs() failed on the spool or log directories while checking their
7363 size for availability, Exim confusingly gave the error "space shortage".
7364 Furthermore, in debugging mode it crashed with a floating point exception.
7365 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7366 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7367 serious problem, Exim now writes to the main and panic logs when this
7368 happens, with details of the failure. It then refuses to accept the
7369 incoming message, giving the message "spool directory problem" or "log
7370 directory problem" with a 421 code for SMTP messages.
7372 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7373 0, 1, and 2 exist, because some OS complain for execs without them (see
7374 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7375 descriptors. However, the code omitted to check that the open succeeded,
7376 causing mysterious errors if for some reason the permissions on /dev/null
7377 got screwed. Now Exim writes a message to the main and panic logs, and
7378 bombs out if it can't open /dev/null.
7380 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7381 interact so that it is all more flexible. It is supposed to remain
7382 backwards compatible. Also added extra_local_interfaces.
7384 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7385 to bomb out with an assertion failure - to the client this appears as a
7386 connection drop. This problem occurs in the part of the code that was taken
7387 from the Samba project. Fortunately, the assertion is in a very simple
7388 function, so I have fixed this by reproducing the function inline in the
7389 one place where it is called, and arranging for authentication to fail
7390 instead of killing the process with assert().
7392 75. The SPA client code was not working when the server requested OEM rather
7393 than Unicode encoding.
7395 76. Added code to make require_files with a specific uid setting more usable in
7396 the case where statting the file as root fails - usually a non-root-mounted
7397 NFS file system. When this happens and the failure is EACCES, Exim now
7398 forks a subprocess and does the per-uid checking as the relevant uid.
7400 77. Added process_log_path.
7402 78. If log_file_path was not explicitly set, a setting of check_log_space or
7403 check_log_inodes was ignored.
7405 79. If a space check for the spool or log partitions fails, the incident is now
7406 logged. Of course, in the latter case the data may get lost...
7408 80. Added the %p formatting code to string_format() so that it can be used to
7409 print addresses in debug_print(). Adjusted all the address printing in the
7410 debugging in store.c to use %p rather than %d.
7412 81. There was a concern that a line of code in smtp_in.c could overflow a
7413 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7414 initially expressed, the concern was not well-founded, because trailing
7415 spaces are removed early. However, if the trailing spaces were followed by
7416 a NULL, they did not get removed, so the overflow was possible. Two fixes
7419 (a) I re-wrote the offending code in a cleaner fashion.
7420 (b) If an incoming SMTP command contains a NULL character, it is rejected
7423 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7424 runs initgroups(), so that if the Exim user is in any additional groups,
7425 they will be used during message reception.
7431 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob