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.
10 HS/01 Handle trailing backslash gracefully. (CVE-2019-15846)
16 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
22 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
23 buffer overrun for (non-chunking) other transports.
25 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
26 TLS1.3, means that a server rejecting a client certificate is not visible
27 to the client until the first read of encrypted data (typically the
28 response to EHLO). Add detection for that case and treat it as a failed
29 TLS connection attempt, so that the normal retry-in-clear can work (if
32 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
33 and/or domain. Found and fixed by Jason Betts.
35 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
36 configuration). If a CNAME target was not a wellformed name pattern, a
39 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
40 the OS reports them interleaved with other addresses.
42 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
43 used both for input and for a verify callout, both encrypted, SMTP
44 responses being sent by the server could be lost. This resulted in
45 dropped connections and sometimes bounces generated by a peer sending
48 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
49 string. Previously it was possible to cause undefined behaviour in a
50 library routine (usually a crash). Found by "zerons".
52 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
55 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
56 API was removed, so update to use the newer ones.
59 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
60 to changes apparently associated with TLS1.3 handling some of the APIs
61 previously used were either nonfunctional or inappropriate. Strings
62 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
63 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
64 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
65 This affects log line X= elements, the $tls_{in,out}_cipher variables,
66 and the use of specific cipher names in the encrypted= ACL condition.
72 JH/01 Remove code calling the customisable local_scan function, unless a new
73 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
75 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
76 non-signal-safe functions being used.
78 JH/03 Bug 2269: When presented with a received message having a stupidly large
79 number of DKIM-Signature headers, disable DKIM verification to avoid
80 a resource-consumption attack. The limit is set at twenty.
82 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
83 report of oldest_pass in ${authres } in consequence, and separate out
84 some descriptions of reasons for verification fail.
86 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
87 files in the spool were present and unlocked. A queue-runner could spot
88 them, resulting in a duplicate delivery. Fix that by doing the unlock
89 after the unlink. Investigation by Tim Stewart. Take the opportunity to
90 add more error-checking on spoolfile handling while that code is being
93 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
94 No known attacks, no CVE, this is defensive hardening.
96 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
97 a queue-runner could start a delivery while other operations were ongoing.
98 Cutthrough delivery was a common victim, resulting in duplicate delivery.
99 Found and investigated by Tim Stewart. Fix by using the open message data
100 file handle rather than opening another, and not locally closing it (which
101 releases a lock) for that case, while creating the temporary .eml format
102 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
104 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
105 $sender_verify_failure/$recipient_verify_failure to "random".
107 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
110 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
111 Previously this would segfault.
113 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
116 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
117 like zero, since the resolver should be doing this for us, But we need one
118 as a CNAME but no MX presence gets the CNAME returned; we need to check
119 that doesn't point to an MX to declare it "no MX returned" rather than
120 "error, loop". A new main option is added so the older capability of
121 following some limited number of chain links is maintained.
123 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
125 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
126 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
127 now-deprecated earlier definitions used only specified the range up to TLS
128 1.2 (in the older-version library docs).
130 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
132 JH/15 Rework TLS client-side context management. Stop using a global, and
133 explicitly pass a context around. This enables future use of TLS for
134 connections to service-daemons (eg. malware scanning) while a client smtp
135 connection is using TLS; with cutthrough connections this is quite likely.
137 JH/16 Fix ARC verification to do AS checks in reverse order.
139 JH/17 Support a "tls" option on the ${readsocket } expansion item.
141 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
142 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
143 Previously the "utf8" would be re-prepended for every additional message.
145 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
146 Previously thery were accepted, resulting in issues when attempting to
147 forward messages to a non-supporting MTA.
149 PP/02 Let -n work with printing macros too, not just options.
151 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
152 one parent address was copied, and bogus data was used at delivery-logging
153 time. Either a crash (after delivery) or bogus log data could result.
154 Discovery and analysis by Tim Stewart.
156 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
157 Previously if the string ended mid-character, we did not insert the
158 promised '?' replacement.
160 PP/04 Documentation: current string operators work on bytes, not codepoints.
162 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
163 should pack well giving a smaller memory footprint so better caching and
164 therefore performance. Group the declarations where this can't be done so
165 that the byte-sized flag variables are not interspersed among pointer
166 variables, giving a better chance of good packing by the compiler.
168 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
169 non-null, to avoid issues with sites running BATV. Previously reports were
170 sent with an empty envelope sender so looked like bounces.
172 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
173 The ignore_error flag wasn't being returned from the filter subprocess so
174 was not set for later routers. Investigation and fix by Matthias Kurz.
176 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
177 and a msg:complete for the whole, when a message is manually removed using
178 -Mrm. Developement by Matthias Kurz, hacked on by JH.
180 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
181 a "Gnu special" function, asprintf() in the DB utility binary builds; I
182 hope that is portable enough.
184 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
185 requiring a known-CA anchor certificate; make it now rely entirely on the
186 TLSA as an anchor. Checking the name on the leaf cert against the name
187 on the A-record for the host is still done for TA (but not for EE mode).
189 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
190 included in delivery lines for non-proxied connections, when compiled with
191 SUPPORT_SOCKS and running with proxy logging enabled.
193 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
194 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
195 move the existing event to fire before the normal logging of message
196 failure so that custom logging is bracketed by normal logging.
198 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
199 msg:fail:internal event. Developement by Matthias Kurz.
201 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
202 far too small for todays use of crypto signatures stored there. Go all
203 the way to the max DNS message size of 64kB, even though this might be
204 overmuch for IOT constrained device use.
206 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
207 copy a string over itself. The library routine is documented as not
208 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
210 JH/32 For main options check_spool_space and check_inode_space, where the
211 platform supports 64b integers, support more than the previous 2^31 kB
212 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
213 the previous G, M, k.
215 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
216 $authenticated_fail_id variable on authentication failure. Previously
219 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
220 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
221 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
222 GNUTLS_SEC_PARAM_MEDIUM.
224 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
225 side. Previously we would continue as if no SNI had been received.
227 JH/36 Harden the handling of string-lists. When a list consisted of a sole
228 "<" character, which should be a list-separator specification, we walked
229 off past the nul-terimation.
231 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
232 causes) even when the retry time is not yet met. Previously they were
233 not, meaning that when (say) an account was over-quota and temp-rejecting,
234 and multiple senders' messages were queued, only one sender would get
235 notified on each configured delay_warning cycle.
237 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
239 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
240 cipher-suites, an error can be left on the stack even for a succeeding
241 accept; this results in impossible error messages when a later operation
244 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
245 return error codes indicating retry. Under TLS1.3 this becomes required.
247 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
248 it only wrote the new authenticators, resulting in a lack of tracking of
249 peer changes of ESMTP extensions until the next cache flush.
251 JH/41 Fix the loop reading a message header line to check for integer overflow,
252 and more-often against header_maxsize. Previously a crafted message could
253 induce a crash of the recive process; now the message is cleanly rejected.
255 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
256 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
262 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
263 When redis_servers is set to a list of > 1 element, and the Redis servers
264 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
265 case of MOVED into a DEFER case instead, thus moving the query onto the
266 next server in the list. For a cluster of N elements, all N servers must
267 be defined in redis_servers.
269 GF/02 Catch and remove uninitialized value warning in exiqsumm
270 Check for existence of @ARGV before looking at $ARGV[0]
272 JH/01 Replace the store_release() internal interface with store_newblock(),
273 which internalises the check required to safely use the old one, plus
274 the allocate and data copy operations duplicated in both (!) of the
275 extant use locations.
277 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
278 modifier. This matches the restriction on the commandline.
280 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
281 Previously only the last row was returned.
283 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
284 we assumed that tags in the header were well-formed, and parsed the
285 element content after inspecting only the first char of the tag.
286 Assumptions at that stage could crash the receive process on malformed
289 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
290 While running the DKIM ACL we operate on the Permanent memory pool so that
291 variables created with "set" persist to the DATA ACL. Also (at any time)
292 DNS lookups that fail create cache records using the Permanent pool. But
293 expansions release any allocations made on the current pool - so a dnsdb
294 lookup expansion done in the DKIM ACL releases the memory used for the
295 DNS negative-cache, and bad things result. Solution is to switch to the
296 Main pool for expansions.
297 While we're in that code, add checks on the DNS cache during store_reset,
298 active in the testsuite.
299 Problem spotted, and debugging aided, by Wolfgang Breyha.
301 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
302 When none of the hosts presented to a transport match an already-open
303 connection, close it and proceed with the list. Previously we would
304 queue the message. Spotted by Lena with Yahoo, probably involving
307 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
308 Previously a spurious "250 OK id=" response was appended to the proper
311 JH/08 The "support for" informational output now, which built with Content
312 Scanning support, has a line for the malware scanner interfaces compiled
313 in. Interface can be individually included or not at build time.
315 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
316 by the template makefile "src/EDITME". The "STREAM" support for an older
317 ClamAV interface method is removed.
319 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
320 rows affected is given instead).
322 JH/11 The runtime Berkeley DB library version is now additionally output by
323 "exim -d -bV". Previously only the compile-time version was shown.
325 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
326 SMTP connection. Previously, when one had more receipients than the
327 first, an abortive onward connection was made. Move to full support for
328 multiple onward connections in sequence, handling cutthrough connection
329 for all multi-message initiating connections.
331 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
332 routers. Previously, a multi-recipient message would fail to match the
333 onward-connection opened for the first recipient, and cause its closure.
335 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
336 a timeout on read on a GnuTLS initiating connection, resulting in the
337 initiating connection being dropped. This mattered most when the callout
338 was marked defer_ok. Fix to keep the two timeout-detection methods
341 JH/15 Relax results from ACL control request to enable cutthrough, in
342 unsupported situations, from error to silently (except under debug)
343 ignoring. This covers use with PRDR, frozen messages, queue-only and
346 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
348 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
349 metadata, resulting in a crash in free().
351 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
352 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
353 Broken also in d185889f4, with init system revamp.
355 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
356 Previously we abruptly closed the connection after reading a malware-
357 found indication; now we go on to read the "scan ok" response line,
360 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
361 ACL. Previously, a crash would result.
363 JH/19 Speed up macro lookups during configuration file read, by skipping non-
364 macro text after a replacement (previously it was only once per line) and
365 by skipping builtin macros when searching for an uppercase lead character.
367 JH/20 DANE support moved from Experimental to mainline. The Makefile control
368 for the build is renamed.
370 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
371 was allocated for every new TLS startup, meaning one per message. Fix
372 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
374 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
375 reported the original. Fix to report (as far as possible) the ACL
376 result replacing the original.
378 JH/23 Fix memory leak during multi-message connections using STARTTLS under
379 OpenSSL. Certificate information is loaded for every new TLS startup,
380 and the resources needed to be freed.
382 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
384 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
387 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
388 DATA response info to the (existing) per-recipient response info for
389 the "C=" log element. It can have useful tracking info from the
390 destination system. Patch from Simon Arlott.
392 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
393 length value. Previously this would segfault.
395 HS/02 Support Avast multiline protoocol, this allows passing flags to
396 newer versions of the scanner.
398 JH/28 Ensure that variables possibly set during message acceptance are marked
399 dead before release of memory in the daemon loop. This stops complaints
400 about them when the debug_store option is enabled. Discovered specifically
401 for sender_rate_period, but applies to a whole set of variables.
402 Do the same for the queue-runner and queue-list loops, for variables set
403 from spool message files. Do the same for the SMTP per-message loop, for
404 certain variables indirectly set in ACL operations.
406 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
407 as a multi-recipient message from a mailinglist manager). The coding had
408 an arbitrary cutoff number of characters while checking for more input;
409 enforced by writing a NUL into the buffer. This corrupted long / fast
410 input. The problem was exposed more widely when more pipelineing of SMTP
411 responses was introduced, and one Exim system was feeding another.
412 The symptom is log complaints of SMTP syntax error (NUL chars) on the
413 receiving system, and refused recipients seen by the sending system
414 (propating to people being dropped from mailing lists).
415 Discovered and pinpointed by David Carter.
417 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
418 replaced by the ${authresults } expansion.
420 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
422 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
423 allows proper process termination in container environments.
425 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
426 Previously the "final dot" had a newline after it; ensure it is CR,LF.
428 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
429 and "err_perm", deprecated since 4.83 when the RFC-defined words
430 "temperror" and "permerror" were introduced.
432 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
433 transport-filters or DKIM-signing. The restriction was lost in the
434 consolidation of verify-callout and delivery SMTP handling.
435 Extend the restriction to also cover ARC-signing.
437 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
438 in defer=pass mode supply a 450 to the initiator. Previously the message
441 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
442 tls_require_ciphers is used as before.
444 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
445 "pass_unscanned". Only tmpfails from the scanner are written to
446 the paniclog, as they may require admin intervention (permission
447 denied, license issues). Other scanner errors (like decompression
448 bombs) do not cause a paniclog entry.
450 JH/36 Fix reinitialisation of DKIM logging variable between messages.
451 Previously it was possible to log spurious information in receive log
454 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
455 triggered odd behaviour from Outlook Express clients.
457 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
460 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
461 since the IETF WG has not yet settled on that versus the original
462 "bare" representation.
464 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
465 Previously the millisecond value corrupted the output.
466 Fix also for syslog_pid=no and log_selector +pid, for which the pid
467 corrupted the output.
473 JH/01 Rework error string handling in TLS interface so that the caller in
474 more cases is responsible for logging. This permits library-sourced
475 string to be attached to addresses during delivery, and collapses
476 pairs of long lines into single ones.
478 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
479 during configuration. Wildcards are allowed and expanded.
481 JH/02 Rework error string handling in DKIM to pass more info back to callers.
482 This permits better logging.
484 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
485 do not close it down and have the child transport start it up again on
486 the passed-on TCP connection. Instead, proxy the child (and any
487 subsequent ones) for TLS via a unix-domain socket channel. Logging is
488 affected: the continued delivery log lines do not have any DNSSEC, TLS
489 Certificate or OCSP information. TLS cipher information is still logged.
491 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
492 identical IP addresses on different listening ports. Will also affect
495 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
496 add noisy ifdef guards to special-case this sillyness.
497 Patch from Bernd Kuhls.
499 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
500 than 255 are no longer allowed.
502 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
503 Disable the session-cache too, which might reduce our load. Since we
504 currrectly use a new context for every connection, both as server and
505 client, there is no benefit for these.
506 GnuTLS appears to not support tickets server-side by default (we don't
507 call gnutls_session_ticket_enable_server()) but client side is enabled
508 by default on recent versions (3.1.3 +) unless the PFS priority string
511 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
512 <https://reproducible-builds.org/specs/source-date-epoch/>.
514 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
515 the check for any unsuccessful recipients did not notice the limit, and
516 erroneously found still-pending ones.
518 JH/08 Pipeline CHUNKING command and data together, on kernels that support
519 MSG_MORE. Only in-clear (not on TLS connections).
521 JH/09 Avoid using a temporary file during transport using dkim. Unless a
522 transport-filter is involved we can buffer the headers in memory for
523 creating the signature, and read the spool data file once for the
524 signature and again for transmission.
526 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
527 4.77 as the kernel support then wasn't solid, having issues in 64bit
528 mode. Now, it's been long enough. Add support for FreeBSD also.
530 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
531 case where the routing stage had gathered several addresses to send to
532 a host before calling the transport for the first, we previously failed
533 to close down TLS in the old transport process before passing the TCP
534 connection to the new process. The new one sent a STARTTLS command
535 which naturally failed, giving a failed delivery and bloating the retry
536 database. Investigation and fix prototype from Wolfgang Breyha.
538 JH/12 Fix check on SMTP command input synchronisation. Previously there were
539 false-negatives in the check that the sender had not preempted a response
540 or prompt from Exim (running as a server), due to that code's lack of
541 awareness of the SMTP input buffering.
543 PP/04 Add commandline_checks_require_admin option.
544 Exim drops privileges sanely, various checks such as -be aren't a
545 security problem, as long as you trust local users with access to their
546 own account. When invoked by services which pass untrusted data to
547 Exim, this might be an issue. Set this option in main configuration
548 AND make fixes to the calling application, such as using `--` to stop
551 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
552 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
553 responses to those, into a single TLS record each way (this usually means
554 a single packet). As a side issue, smtp_enforce_sync now works on TLS
557 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
558 affects you only if you're dancing at the edge of the param size limits.
559 If you are, and this message makes sense to you, then: raise the
560 configured limit or use OpenSSL 1.1. Nothing we can do for older
563 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
564 cmdline element to get the documented default one. Previously it was
567 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
568 are unsuable for non-message-specific reasons - eg. connection timeout,
569 banner-time rejection.
571 JH/16 Fix logging of delivery remote port, when specified by router, under
574 PP/06 Repair manualroute's ability to take options in any order, even if one
575 is the name of a transport.
578 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
580 JH/17 Change the list-building routines interface to use the expanding-string
581 triplet model, for better allocation and copying behaviour.
583 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
584 Previously it was constructed the first time a possibly-matching string
585 was met in the configuration file input during startup; now it is done
588 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
589 compatible one, to avoid the (poorly documented) possibility of a config
590 file in the working directory redirecting the DB files, possibly correpting
591 some existing file. CVE-2017-10140 assigned for BDB.
593 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
594 cache-hot. Previously, although the result was properly cached, the
595 initial verify call returned a defer.
597 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
598 the main verify for receipient in uncached-mode.
600 JH/22 Retire historical build files to an "unsupported" subdir. These are
601 defined as "ones for which we have no current evidence of testing".
603 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
604 if present. Previously it was ignored.
606 JH/24 Start using specified-initialisers in C structure init coding. This is
607 a C99 feature (it's 2017, so now considered safe).
609 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
610 if was a fixed-sized field and bitmask ops via macros; it is now more
613 PP/07 GitHub PR 56: Apply MariaDB build fix.
614 Patch provided by Jaroslav Škarvada.
616 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
617 during Coverity cleanups [4.87 JH/47]
618 Diagnosis and fix provided by Michael Fischer v. Mollard.
620 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
621 the right size to place the terminating semicolon on its own folded
622 line, the header hash was calculated to an incorrect value thanks to
623 the (relaxed) space the fold became.
625 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
626 and confused the parent.
628 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
629 which could crash as a result. This could lead to undeliverable messages.
631 JH/28 Logging: "next input sent too soon" now shows where input was truncated
634 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
635 matters on fast-turnover and PID-randomising systems, which were getting
636 out-of-order delivery.
638 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
639 a possibly-overlapping copy. The symptom was that "Remote host closed
640 connection in response to HELO" was logged instead of the actual 4xx
643 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
644 Previously only that bufferd was discarded, resulting in SYMTP command
647 JH/32 DKIM: when a message has multiple signatures matching an identity given
648 in dkim_verify_signers, run the dkim acl once for each. Previously only
649 one run was done. Bug 2189.
651 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
652 "panic the current process" to "deliberately defer". The panic log is
653 still written with the problem list name; the mail and reject logs now
654 get a temp-reject line for the message that was being handled, saying
655 something like "domains check lookup or other defer". The SMTP 451
656 message is still "Temporary local problem".
658 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
659 A crafted sequence of BDAT commands could result in in-use memory beeing
660 freed. CVE-2017-16943.
662 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
663 from SMTP input. Previously it was always done; now only done for DATA
664 and not BDAT commands. CVE-2017-16944.
666 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
667 to the message (such as an overlong header line). Previously this was
668 not done and we did not exit BDAT mode. Followon from the previous item
669 though a different problem.
675 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
676 than -2003 did; needs libidn2 in addition to libidn.
678 JH/02 The path option on a pipe transport is now expanded before use.
680 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
681 Patch provided by "Björn", documentation fix added too.
683 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
684 missing a wire-to-host endian conversion.
686 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
687 close after a BDAT command line could be taken as a following command,
688 giving a synch failure. Fix by only checking for synch immediately
689 before acknowledging the chunk.
691 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
692 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
693 macro. Patches provided by Josh Soref.
695 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
696 Previously we did not; the RFC seems ambiguous and VRFY is not listed
697 by IANA as a service extension. However, John Klensin suggests that we
700 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
701 the dkim code may be unix-mode line endings rather than smtp wire-format
702 CRLF, so prepend a CR to any bare LF.
704 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
705 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
707 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
708 the first were themselves being wrongly included in the feed into dkim
709 processing; with most chunk sizes in use this resulted in an incorrect
710 body hash calculated value.
712 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
713 DKIM signature block, for verification. Although advised against by
714 standards it is specifically not ruled illegal.
716 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
718 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
719 missing a body hash (the bh= tag).
721 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
722 It seems that HAProxy sends the Proxy Protocol information in clear and
723 only then does a TLS startup, so do the same.
725 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
726 TCP connections (such as for Spamd) unless the daemon successfully set
727 Fast Open mode on its listening sockets. This fixes breakage seen on
728 too-old kernels or those not configured for Fast Open, at the cost of
729 requiring both directions being enabled for TFO, and TFO never being used
730 by non-daemon-related Exim processes.
732 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
733 endings, at least on the first header line. Try to canonify any that get
734 past that check, despite the cost.
736 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
737 now limited to an arbitrary five deep, while parsing addresses with the
738 strip_excess_angle_brackets option enabled.
740 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
741 instead leave the unprompted TLS handshake in socket buffer for the
742 TLS library to consume.
744 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
746 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
748 JH/16 Drop variables when they go out of scope. Memory management drops a whole
749 region in one operation, for speed, and this leaves assigned pointers
750 dangling. Add checks run only under the testsuite which checks all
751 variables at a store-reset and panics on a dangling pointer; add code
752 explicitly nulling out all the variables discovered. Fixes one known
753 bug: a transport crash, where a dangling pointer for $sending_ip_address
754 originally assigned in a verify callout, is re-used.
756 PP/06 Drop '.' from @INC in various Perl scripts.
758 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
760 PP/08 Reduce a number of compilation warnings under clang; building with
761 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
762 should be warning-free.
764 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
766 HS/01 Fix portability problems introduced by PP/08 for platforms where
767 realloc(NULL) is not equivalent to malloc() [SunOS et al].
769 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
770 chunk. This allows us to accept broken chunked messages. We need a more
771 general solution here.
773 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
774 already-broken messages in the queue.
776 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
778 JH/19 Fix reference counting bug in routing-generated-address tracking.
784 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
785 supports it and a size is available (ie. the sending peer gave us one).
787 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
788 years of being deprecated). The replacements are the ACLs
789 acl_smtp_mime and acl_not_smtp_mime.
791 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
792 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
793 or even an in-clear connection were permitted. Now, if the host lookup
794 was dnssec and dane was requested then the host is only used if the
795 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
796 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
797 if one fails this test.
798 This means that a poorly-configured remote DNS will make it incommunicado;
799 but it protects against a DNS-interception attack on it.
801 JH/04 Bug 1810: make continued-use of an open smtp transport connection
802 non-noisy when a race steals the message being considered.
804 JH/05 If main configuration option tls_certificate is unset, generate a
805 self-signed certificate for inbound TLS connections.
807 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
808 in rewrites and routers.
810 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
811 and logged a warning sing 4.83; now they are a configuration file error.
813 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
814 (lacking @domain). Apply the same qualification processing as RCPT.
816 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
818 JH/10 Support ${sha256:} applied to a string (as well as the previous
821 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
822 a cutthrough deliver is pending, as we always want to make a connection.
823 This also avoids re-routing the message when later placing the cutthrough
824 connection after a verify cache hit.
825 Do not update it with the verify result either.
827 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
828 when routing results in more than one destination address.
830 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
831 signing (which inhibits the cutthrough capability). Previously only
832 the presence of an option was tested; now an expansion evaluating as
833 empty is permissible (obviously it should depend only on data available
834 when the cutthrough connection is made).
836 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
837 the relevant preceding SMTP command did not note the pipelining mode.
839 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
840 Previously they were not counted.
842 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
843 as one having no matching records. Previously we deferred the message
844 that needed the lookup.
846 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
847 distinguished as "(=".
849 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
850 for missing MX records. Previously it only worked for missing A records.
852 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
854 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
855 after the data-go-ahead and data-ack. Patch from Jason Betts.
857 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
858 even for a "none" policy. Patch from Tony Meyer.
860 JH/22 Fix continued use of a connection for further deliveries. If a port was
861 specified by a router, it must also match for the delivery to be
864 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
865 When one of the recipients of a message was unsuitable for the connection
866 (has no matching addresses), we lost track of needing to mark it
867 deferred. As a result mail would be lost.
869 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
871 JH/25 Decoding ACL controls is now done using a binary search; the source code
872 takes up less space and should be simpler to maintain. Merge the ACL
873 condition decode tables also, with similar effect.
875 JH/26 Fix problem with one_time used on a redirect router which returned the
876 parent address unchanged. A retry would see the parent address marked as
877 delivered, so not attempt the (identical) child. As a result mail would
880 JH/27 Fix a possible security hole, wherein a process operating with the Exim
881 UID can gain a root shell. Credit to http://www.halfdog.net/ for
882 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
885 JH/28 Enable {spool,log} filesystem space and inode checks as default.
886 Main config options check_{log,spool}_{inodes,space} are now
887 100 inodes, 10MB unless set otherwise in the configuration.
889 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
890 Previously it only applied to the main-section connection policy
893 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
895 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
896 by me. Added RFC7919 DH primes as an alternative.
898 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
899 are not in the system include path.
901 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
902 GnuTLS, when a session startup failed (eg because the client disconnected)
903 Exim did stdio operations after fclose. This was exposed by a recent
904 change which nulled out the file handle after the fclose.
906 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
907 signed directly by the cert-signing cert, rather than an intermediate
908 OCSP-signing cert. This is the model used by LetsEncrypt.
910 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
912 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
913 an incoming connection.
915 HS/02 Bug 1802: Do not half-close the connection after sending a request
918 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
919 fallback to "prime256v1".
921 JH/34 SECURITY: Use proper copy of DATA command in error message.
922 Could leak key material. Remotely exploitable. CVE-2016-9963.
928 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
929 and 3.4.4 - once the server is enabled to respond to an OCSP request
930 it does even when not requested, resulting in a stapling non-aware
931 client dropping the TLS connection.
933 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
934 support variable-length bit vectors. No functional change.
936 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
937 The I= interface field on outgoing lines is now after the H= remote
938 host field, same as incoming lines. There is a separate
939 outgoing_interface log selector which allows you to disable the
942 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
943 If not running log_selector +smtp_connection the mainlog would be held
944 open indefinitely after a "too many connections" event, including to a
945 deleted file after a log rotate. Leave the per net connection logging
946 leaving it open for efficiency as that will be quickly detected by the
947 check on the next write.
949 HS/01 Bug 1671: Fix post transport crash.
950 Processing the wait-<transport> messages could crash the delivery
951 process if the message IDs didn't exist for some reason. When
952 using 'split_spool_directory=yes' the construction of the spool
953 file name failed already, exposing the same netto behaviour.
955 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
956 mime_regex ACL conditions.
958 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
959 to DSN fail messages (bounces): remote IP, remote greeting, remote response
960 to HELO, local diagnostic string.
962 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
963 log line to debug. Even when configured with a tls authenticator many
964 client connections are expected to not authenticate in this way, so
965 an authenticate fail is not an error.
967 HS/02 Add the Exim version string to the process info. This way exiwhat
968 gives some more detail about the running daemon.
970 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
971 matter for fast-change records such as DNSBLs.
973 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
974 retry record, even if constant. There may be multiple transports with
975 different interface settings and the retry behaviour needs to be kept
978 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
980 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
982 JH/10 Bug 840: fix log_defer_output option of pipe transport
984 JH/11 Bug 830: use same host for all RCPTS of a message, even under
985 hosts_randomize. This matters a lot when combined with mua_wrapper.
987 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
988 ${quote_pgsql:<string>} operator.
990 JH/13 Bug 1708: avoid misaligned access in cached lookup.
992 JH/14 Change header file name for freeradius-client. Relevant if compiling
993 with Radius support; from the Gentoo tree and checked under Fedora.
995 JH/15 Bug 1712: Introduce $prdr_requested flag variable
997 JH/16 Bug 1714: Permit an empty string as expansion result for transport
998 option transport_filter, meaning no filtering.
1000 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1002 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1003 defaults to "*" (all hosts). The variable is now available when not built
1004 with TLS, default unset, mainly to enable keeping the testsuite sane.
1005 If a server certificate is not supplied (via tls_certificate) an error is
1006 logged, and clients will find TLS connections fail on startup. Presumably
1007 they will retry in-clear.
1008 Packagers of Exim are strongly encouraged to create a server certificate
1009 at installation time.
1011 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1012 with the $config_file variable.
1014 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1015 in transport context, after the attempt, and per-recipient. The latter type
1016 is per host attempted. The event data is the error message, and the errno
1017 information encodes the lookup type (A vs. MX) used for the (first) host,
1018 and the trailing two digits of the smtp 4xx response.
1020 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1021 to write to mainlog (or rejectlog, paniclog) in the window between file
1022 creation and permissions/ownership being changed. Particularly affects
1023 installations where exicyclog is run as root, rather than exim user;
1024 result is that the running daemon panics and dies.
1026 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1028 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1029 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1030 "pri" and "weight". Note that the previous implicit priority given by the
1031 list order is no longer honoured.
1033 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1034 for DKIM processing.
1036 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1037 by defining SUPPORT_SOCKS.
1039 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1040 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1041 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1042 variables are renamed to proxy_{local,external}_{address,port}.
1044 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1045 for a build by defining SUPPORT_I18N
1047 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1048 of the query string, and make ${quote_redis:} do that quoting.
1050 JH/29 Move Events support from Experimental to mainline, enabled by default
1051 and removable for a build by defining DISABLE_EVENT.
1053 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1055 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1056 cached by the daemon.
1058 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1059 by defining LOOKUP_REDIS. The libhiredis library is required.
1061 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1062 keys are given for lookup.
1064 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1065 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1066 only supported when built with TLS support. The PolarSSL SHA routines
1067 are still used when the TLS library is too old for convenient support.
1069 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1070 openssl_options), for security. OpenSSL forces this from version 1.1.0
1071 server-side so match that on older versions.
1073 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1074 allocation for $value could be released as the expansion processing
1075 concluded, but leaving the global pointer active for it.
1077 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1078 and to use the domains and local_parts ACL conditions.
1080 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1081 incorrectly not doubled on cutthrough transmission, hence seen as a
1082 body-termination at the receiving system - resulting in truncated mails.
1083 Commonly the sender saw a TCP-level error, and retransmitted the message
1084 via the normal store-and-forward channel. This could result in duplicates
1085 received - but deduplicating mailstores were liable to retain only the
1086 initial truncated version.
1088 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1090 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1092 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1093 we're in there, support oversigning also; bug 1309.
1095 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1097 HS/04 Add support for keep_environment and add_environment options.
1099 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1100 either intentional arithmetic overflow during PRNG, or testing config-
1103 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1104 delivery resulted in actual delivery. Cancel cutthrough before DATA
1107 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1108 rejects a recipient: pass the reject to the originator.
1110 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1111 Many were false-positives and ignorable, but it's worth fixing the
1114 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1115 for the new environment-manipulation done at startup. Move the routines
1116 from being local to tls.c to being global via the os.c file.
1118 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1119 an extract embedded as result-arg for a map, the first arg for extract
1120 is unavailable so we cannot tell if this is a numbered or keyed
1121 extraction. Accept either.
1127 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1130 JH/02 The smtp transport option "multi_domain" is now expanded.
1132 JH/03 The smtp transport now requests PRDR by default, if the server offers
1135 JH/04 Certificate name checking on server certificates, when exim is a client,
1136 is now done by default. The transport option tls_verify_cert_hostnames
1137 can be used to disable this per-host. The build option
1138 EXPERIMENTAL_CERTNAMES is withdrawn.
1140 JH/05 The value of the tls_verify_certificates smtp transport and main options
1141 default to the word "system" to access the system default CA bundle.
1142 For GnuTLS, only version 3.0.20 or later.
1144 JH/06 Verification of the server certificate for a TLS connection is now tried
1145 (but not required) by default. The verification status is now logged by
1146 default, for both outbound TLS and client-certificate supplying inbound
1149 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1152 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1153 Status Notification (bounce) messages are now MIME format per RFC 3464.
1154 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1155 under the control of the dsn_advertise_hosts option, and routers may
1156 have a dsn_lasthop option.
1158 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1159 default, modifiable by a malware= option. The list separator for
1160 the options can now be changed in the usual way. Bug 68.
1162 JH/10 The smtp_receive_timeout main option is now expanded before use.
1164 JH/11 The incoming_interface log option now also enables logging of the
1165 local interface on delivery outgoing connections.
1167 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1168 if the interface and destination host and port all match.
1170 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1173 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1174 Patch from Andrew Lewis.
1176 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1177 now supports optional time-restrictions, weighting, and priority
1178 modifiers per server. Patch originally by <rommer@active.by>.
1180 JH/16 The spamd_address main option now supports a mixed list of local
1181 and remote servers. Remote servers can be IPv6 addresses, and
1182 specify a port-range.
1184 JH/17 Bug 68: The spamd_address main option now supports an optional
1185 timeout value per server.
1187 JH/18 Bug 1581: Router and transport options headers_add/remove can
1188 now have the list separator specified.
1190 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1193 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1196 JH/21 Support for the A6 type of dns record is withdrawn.
1198 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1199 rather than the verbs used.
1201 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1202 from 255 to 1024 chars.
1204 JH/24 Verification callouts now attempt to use TLS by default.
1206 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1207 are generic router options now. The defaults didn't change.
1209 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1210 Original patch from Alexander Shikoff, worked over by JH.
1212 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1213 files if ZCAT_COMMAND is not executable.
1215 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1217 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1219 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1220 Normally benign, it bites when the pair was led to by a CNAME;
1221 modern usage is to not canonicalize the domain to a CNAME target
1222 (and we were inconsistent anyway for A-only vs AAAA+A).
1224 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1226 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1227 when evaluating $sender_host_dnssec.
1229 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1230 $sender_helo_dnssec variable.
1232 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1234 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1236 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1238 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1239 documented as working, but never had. Support all but $spam_report.
1241 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1242 added for tls authenticator.
1244 HS/03 Add perl_taintmode main config option
1250 TL/01 When running the test suite, the README says that variables such as
1251 no_msglog_check are global and can be placed anywhere in a specific
1252 test's script, however it was observed that placement needed to be near
1253 the beginning for it to behave that way. Changed the runtest perl
1254 script to read through the entire script once to detect and set these
1255 variables, reset to the beginning of the script, and then run through
1256 the script parsing/test process like normal.
1258 TL/02 The BSD's have an arc4random API. One of the functions to induce
1259 adding randomness was arc4random_stir(), but it has been removed in
1260 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1261 function when detected.
1263 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1264 cause callback expansion.
1266 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1267 syntax errors in an expansion can be treated as a string instead of
1268 logging or causing an error, due to the internal use of bool_lax
1269 instead of bool when processing it.
1271 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1272 server certificates when making smtp deliveries.
1274 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1276 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1278 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1280 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1281 Merged patch from Sebastian Wiedenroth.
1283 JH/05 Fix results-pipe from transport process. Several recipients, combined
1284 with certificate use, exposed issues where response data items split
1285 over buffer boundaries were not parsed properly. This eventually
1286 resulted in duplicates being sent. This issue only became common enough
1287 to notice due to the introduction of connection certificate information,
1288 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1290 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1291 size buffer was used, resulting in syntax errors when an expansion
1294 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1295 version 3.3.6 or later.
1297 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1298 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1299 both become "event_action", the variables become $event_name, $event_data
1300 and $event_defer_errno. There is a new variable $verify_mode, usable in
1301 routers, transports and related events. The tls:cert event is now also
1302 raised for inbound connections, if the main configuration event_action
1305 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1306 early OCSP support, but no stapling (appears to be less than 1.0.0).
1308 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1309 server certificate names available under the smtp transport option
1310 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1313 JH/10 Time-related extraction expansions from certificates now use the main
1314 option "timezone" setting for output formatting, and are consistent
1315 between OpenSSL and GnuTLS compilations. Bug 1541.
1317 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1318 encoded parameter in the incoming message. Bug 1558.
1320 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1321 include certificate info, eximon was claiming there were spoolfile
1324 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1326 JH/14 Log delivery-related information more consistently, using the sequence
1327 "H=<name> [<ip>]" wherever possible.
1329 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1330 are problematic for Debian distribution, omit them from the release
1333 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1335 JH/16 Fix string representation of time values on 64bit time_t architectures.
1338 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1339 output list separator was used.
1344 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1345 checkers that were complaining about end of non-void function with no
1348 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1349 This was a regression introduced in 4.83 by another bugfix.
1351 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1353 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1354 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1360 TF/01 Correctly close the server side of TLS when forking for delivery.
1362 When a message was received over SMTP with TLS, Exim failed to clear up
1363 the incoming connection properly after forking off the child process to
1364 deliver the message. In some situations the subsequent outgoing
1365 delivery connection happened to have the same fd number as the incoming
1366 connection previously had. Exim would try to use TLS and fail, logging
1367 a "Bad file descriptor" error.
1369 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1370 utilities have not been installed.
1372 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1373 temporary space as the ACL may create new global variables.
1375 TL/01 LDAP support uses per connection or global context settings, depending
1376 upon the detected version of the libraries at build time.
1378 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1379 to extract and use the src ip:port in logging and expansions as if it
1380 were a direct connection from the outside internet. PPv2 support was
1381 updated based on HAProxy spec change in May 2014.
1383 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1385 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1386 Properly escape header and check for NULL return.
1388 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1391 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1393 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1394 characters in header names, implemented as a verify condition.
1395 Contributed by Michael Fischer v. Mollard.
1397 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1398 results permerror and temperror. Previous values are deprecated but
1399 still accepted. In a future release, err_perm and err_temp will be
1400 completely removed, which will be a backward incompatibility if the
1401 ACL tests for either of these two old results. Patch contributed by
1402 user bes-internal on the mailing list.
1404 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1406 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1407 selectors, in both main and reject logs.
1409 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1412 JH/07 Add malware type "sock" for talking to simple daemon.
1414 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1416 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1417 routers/transports under cutthrough routing.
1419 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1420 numbers. Touch up "bool" conditional to keep the same definition.
1422 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1424 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1426 JH/12 Expand items in router/transport headers_add or headers_remove lists
1427 individually rather than the list as a whole. Bug 1452.
1429 Required for reasonable handling of multiple headers_ options when
1430 they may be empty; requires that headers_remove items with embedded
1431 colons must have them doubled (or the list-separator changed).
1433 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1434 view the policy declared in the DMARC record. Currently, $dmarc_status
1435 is a combined value of both the record presence and the result of the
1438 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1440 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1441 dnslookup router and the smtp transport (applying to the forward
1444 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1445 of ldap servers used for a specific lookup. Patch provided by Heiko
1448 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1449 New variable $lookup_dnssec_authenticated for observability.
1451 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1452 Patch submitted by Lars Timman.
1454 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1456 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1457 Requires trusted mode and valid format message id, aborts otherwise.
1458 Patch contributed by Heiko Schlichting.
1460 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1461 certextract with support for various fields. Bug 1358.
1463 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1464 is requested by default, modifiable by smtp transport option
1467 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1468 operate on certificate variables to give certificate fingerprints
1469 Also new ${sha256:cert_variable}.
1471 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1473 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1474 Christian Aistleitner.
1476 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1478 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1479 file. Patch from Wolfgang Breyha.
1481 JH/25 Expand the coverage of the delivery $host and $host_address to
1482 client authenticators run in verify callout. Bug 1476.
1484 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1485 align with daemon_smtp_ports. Bug 72.
1487 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1488 support and error reporting did not work properly.
1490 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1491 and is readable. Patch from Andrew Colin Kissa.
1493 TL/14 Enhance documentation of ${run expansion and how it parses the
1494 commandline after expansion, particularly in the case when an
1495 unquoted variable expansion results in an empty value.
1497 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1499 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1500 by renaming away from it. Addresses GH issue 15, reported by
1503 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1505 TL/15 SECURITY: prevent double expansion in math comparison functions
1506 (can expand unsanitized data). Not remotely exploitable.
1513 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1515 PP/02 Make -n do something, by making it not do something.
1516 When combined with -bP, the name of an option is not output.
1518 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1521 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1522 $sender_host_name and config options to manage this, and basic check
1525 PP/05 DSCP support for outbound connections and control modifier for inbound.
1527 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1528 (Only plugin which currently uses this is kerberos4, which nobody should
1529 be using, but we should make it available and other future plugins might
1530 conceivably use it, even though it would break NAT; stuff *should* be
1531 using channel bindings instead).
1533 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1534 name; added for Sendmail compatibility; requires admin caller.
1535 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1536 just ignore it); requires trusted caller.
1537 Also parse but ignore: -Ac -Am -X<logfile>
1540 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1542 TL/02 Add +smtp_confirmation as a default logging option.
1544 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1545 Patch by Magnus Holmgren from 2007-02-20.
1547 TL/04 Bugzilla 1281 - Spec typo.
1548 Bugzilla 1283 - Spec typo.
1549 Bugzilla 1290 - Spec grammar fixes.
1551 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1553 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1555 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1556 mailing list by by Dmitry Isaikin.
1558 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1560 JH/02 Support "G" suffix to numbers in ${if comparisons.
1562 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1564 NM/01 Bugzilla 1197 - Spec typo
1565 Bugzilla 1196 - Spec examples corrections
1567 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1569 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1570 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1573 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1574 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1576 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1577 "acl {{name}{arg}...}", and optional args on acl condition
1580 JH/05 Permit multiple router/transport headers_add/remove lines.
1582 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1584 JH/07 Avoid using a waiting database for a single-message-only transport.
1585 Performance patch from Paul Fisher. Bugzilla 1262.
1587 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1590 JH/09 Add $headers_added variable, with content from use of ACL modifier
1591 add_header (but not yet added to the message). Bugzilla 199.
1593 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1594 Pulled from Bugzilla 817 by Wolfgang Breyha.
1596 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1598 (nb: this is the same fix as in Exim 4.80.1)
1600 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1603 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1606 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1608 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1609 advertise SMTP AUTH mechanism to us, instead of a generic
1610 protocol violation error. Also, make Exim more robust to bad
1611 data from the Dovecot auth socket.
1613 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1615 When a queue runner is handling a message, Exim first routes the
1616 recipient addresses, during which it prunes them based on the retry
1617 hints database. After that it attempts to deliver the message to
1618 any remaining recipients. It then updates the hints database using
1621 So if a recipient address works intermittently, it can get repeatedly
1622 deferred at routing time. The retry hints record remains fresh so the
1623 address never reaches the final cutoff time.
1625 This is a fairly common occurrence when a user is bumping up against
1626 their storage quota. Exim had some logic in its local delivery code
1627 to deal with this. However it did not apply to per-recipient defers
1628 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1630 This change adds a proper retry rule check during routing so that the
1631 final cutoff time is checked against the message's age. We only do
1632 this check if there is an address retry record and there is not a
1633 domain retry record; this implies that previous attempts to handle
1634 the address had the retry_use_local_parts option turned on. We use
1635 this as an approximation for the destination being like a local
1636 delivery, as in LMTP.
1638 I suspect this new check makes the old local delivery cutoff check
1639 redundant, but I have not verified this so I left the code in place.
1641 TF/02 Correct gecos expansion when From: is a prefix of the username.
1643 Test 0254 submits a message to Exim with the header
1647 When I ran the test suite under the user fanf2, Exim expanded
1648 the header to contain my full name, whereas it should have added
1649 a Resent-Sender: header. It erroneously treats any prefix of the
1650 username as equal to the username.
1652 This change corrects that bug.
1654 GF/01 DCC debug and logging tidyup
1655 Error conditions log to paniclog rather than rejectlog.
1656 Debug lines prefixed by "DCC: " to remove any ambiguity.
1658 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1660 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1661 Bug spotted by Jeremy Harris; was flawed since initial commit.
1662 Would have resulted in OCSP responses post-SNI triggering an Exim
1663 NULL dereference and crash.
1665 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1667 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1668 Bug detection, analysis and fix by Samuel Thibault.
1669 Bugzilla 1331, Debian bug #698092.
1671 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1673 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1674 Server implementation by Todd Lyons, client by JH.
1675 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1676 config variable "prdr_enable" controls whether the server
1677 advertises the facility. If the client requests PRDR a new
1678 acl_data_smtp_prdr ACL is called once for each recipient, after
1679 the body content is received and before the acl_smtp_data ACL.
1680 The client is controlled by both of: a hosts_try_prdr option
1681 on the smtp transport, and the server advertisement.
1682 Default client logging of deliveries and rejections involving
1683 PRDR are flagged with the string "PRDR".
1685 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1686 fclose(). Diagnosis by Todd Lyons.
1688 PP/17 Update configure.default to handle IPv6 localhost better.
1689 Patch by Alain Williams (plus minor tweaks).
1692 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1693 This is now consistent with GnuTLS, and is now documented: the
1694 previous undocumented portable approach to treating the option as
1695 unset was to force an expansion failure. That still works, and
1696 an empty string is now equivalent.
1698 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1699 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1700 not performing validation itself.
1702 PP/20 Added force_command boolean option to pipe transport.
1703 Patch from Nick Koston, of cPanel Inc.
1705 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1708 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1710 PP/21 Fix eximon continuous updating with timestamped log-files.
1711 Broken in a format-string cleanup in 4.80, missed when I repaired the
1712 other false fix of the same issue.
1713 Report and fix from Heiko Schlichting.
1716 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1717 Report from Prashanth Katuri.
1719 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1720 It's SecureTransport, so affects any MacOS clients which use the
1721 system-integrated TLS libraries, including email clients.
1723 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1724 using a MIME ACL for non-SMTP local injection.
1725 Report and assistance in diagnosis by Warren Baker.
1727 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1729 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1731 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1732 last id that failed so it may be referenced in subsequent ACL's.
1734 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1737 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1738 ldap library initialization, allowing self-signed CA's to be
1739 used. Also properly sets require_cert option later in code by
1740 using NULL (global ldap config) instead of ldap handle (per
1741 session). Bug diagnosis and testing by alxgomz.
1743 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1744 the src/util/ subdirectory.
1746 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1747 renamed to Transport Post Delivery Action by Jeremy Harris, as
1750 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1751 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1752 redis_servers = needs to be configured which will be used by the redis
1753 lookup. Patch from Warren Baker, of The Packet Hub.
1755 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1757 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1758 hostname or reverse DNS when processing a host list. Used suggestions
1759 from multiple comments on this bug.
1761 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1763 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1764 Missed a few lines, added it to make the runtest require no keyboard
1767 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1768 contains upper case chars. Make router use caseful_local_part.
1770 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1771 support when GnuTLS has been built with p11-kit.
1777 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1779 This, or similar/improved, will also be change PP/11 of 4.82.
1785 PP/01 Handle short writes when writing local log-files.
1786 In practice, only affects FreeBSD (8 onwards).
1787 Bugzilla 1053, with thanks to Dmitry Isaikin.
1789 NM/01 Bugzilla 949 - Documentation tweak
1791 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1794 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1796 PP/02 Implemented gsasl authenticator.
1798 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1800 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1801 `pkg-config foo` for cflags/libs.
1803 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1804 with rest of GSASL and with heimdal_gssapi.
1806 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1807 `pkg-config foo` for cflags/libs for the TLS implementation.
1809 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1810 properties get this fed in as external SSF. A number of robustness
1811 and debugging improvements to the cyrus_sasl authenticator.
1813 PP/08 cyrus_sasl server now expands the server_realm option.
1815 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1816 Patch by Jeremy Harris.
1818 PP/10 Added dbmjz lookup type.
1820 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1822 PP/12 MAIL args handles TAB as well as SP, for better interop with
1823 non-compliant senders.
1824 Analysis and variant patch by Todd Lyons.
1826 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1827 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1828 Patch from Dirk Mueller <dmueller@suse.com>
1830 PP/13 tls_peerdn now print-escaped for spool files.
1831 Observed some $tls_peerdn in wild which contained \n, which resulted
1832 in spool file corruption.
1834 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1835 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1836 or write after TLS renegotiation, which otherwise led to messages
1839 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1840 as a tracking header (ie: a signed header comes before the signature).
1841 Patch from Wolfgang Breyha.
1843 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1844 comma-sep list; embedded commas doubled.
1846 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1848 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1850 Report and patch from Dmitry Banschikov.
1852 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1853 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1854 failures. We appear to now support TLS1.1+ with Exim.
1856 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1857 lets Exim select keys and certificates based upon TLS SNI from client.
1858 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1859 before an outbound SMTP session. New log_selector, +tls_sni.
1861 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1862 NULL dereference. Report and patch from Alun Jones.
1864 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1865 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1866 Not seeing resolver debug output on NetBSD, but suspect this is a
1867 resolver implementation change.
1869 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1870 Left warnings. Added "eximon gdb" invocation mode.
1872 PP/21 Defaulting "accept_8bitmime" to true, not false.
1874 PP/22 Added -bw for inetd wait mode support.
1876 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1877 locate the relevant includes and libraries. Made this the default.
1879 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1880 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1882 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1883 This may cause build issues on older platforms.
1885 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1886 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1887 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1888 Added SNI support via GnuTLS too.
1889 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1891 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1893 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1894 Applied second patch from Janne, implementing suggestion to default
1895 multiple-strings-in-record handling to match SPF spec.
1897 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1899 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1900 read-only, out of scope).
1901 Patch from Wolfgang Breyha, report from Stuart Northfield.
1903 PP/29 Fix three issues highlighted by clang analyser static analysis.
1904 Only crash-plausible issue would require the Cambridge-specific
1905 iplookup router and a misconfiguration.
1906 Report from Marcin Mirosław.
1908 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1910 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1911 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1912 As part of this, removing so much warning spew let me fix some minor
1913 real issues in debug logging.
1915 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1916 assignment on my part. Fixed.
1918 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1919 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1920 Janne Snabb (who went above and beyond: thank you).
1922 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1923 string otherwise requires a connection and a bunch more work and it's
1924 relatively easy to get wrong. Should also expose TLS library linkage
1927 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1928 64-bit ${eval} (JH/03).
1930 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1931 GNU libc to support some of the 64-bit stuff, should not lead to
1932 conflicts. Defined before os.h is pulled in, so if a given platform
1933 needs to override this, it can.
1935 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1936 protection layer was required, which is not implemented.
1937 Bugzilla 1254, patch from Wolfgang Breyha.
1939 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1940 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1941 tls_dhparam take prime identifiers. Also unbreak combination of
1942 OpenSSL+DH_params+TLSSNI.
1944 PP/39 Disable SSLv2 by default in OpenSSL support.
1950 PP/01 Solaris build fix for Oracle's LDAP libraries.
1951 Bugzilla 1109, patch from Stephen Usher.
1953 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1955 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1958 TF/02 Fix a couple more cases where we did not log the error message
1959 when unlink() failed. See also change 4.74-TF/03.
1961 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1962 lock up or crash if it happened to be inside a call to libc when it
1963 got a SIGUSR1 from exiwhat.
1965 The SIGUSR1 handler appends the current process status to the process
1966 log which is later printed by exiwhat. It used to use the general
1967 purpose logging code to do this, but several functions it calls are
1968 not safe for signals.
1970 The new output code in the SIGUSR1 handler is specific to the process
1971 log, and simple enough that it's easy to inspect for signal safety.
1972 Removing some special cases also simplifies the general logging code.
1973 Removing the spurious timestamps from the process log simplifies
1976 TF/04 Improved ratelimit ACL condition.
1978 The /noupdate option has been deprecated in favour of /readonly which
1979 has clearer semantics. The /leaky, /strict, and /readonly update modes
1980 are mutually exclusive. The update mode is no longer included in the
1981 database key; it just determines when the database is updated. (This
1982 means that when you upgrade Exim will forget old rate measurements.)
1984 Exim now checks that the per_* options are used with an update mode that
1985 makes sense for the current ACL. For example, when Exim is processing a
1986 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1987 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1988 must specify per_mail/readonly. If you omit the update mode it defaults to
1989 /leaky where that makes sense (as before) or /readonly where required.
1991 The /noupdate option is now undocumented but still supported for
1992 backwards compatibility. It is equivalent to /readonly except that in
1993 ACLs where /readonly is required you may specify /leaky/noupdate or
1994 /strict/noupdate which are treated the same as /readonly.
1996 A useful new feature is the /count= option. This is a generalization
1997 of the per_byte option, so that you can measure the throughput of other
1998 aggregate values. For example, the per_byte option is now equivalent
1999 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2001 The per_rcpt option has been generalized using the /count= mechanism
2002 (though it's more complicated than the per_byte equivalence). When it is
2003 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2004 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2005 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2006 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2007 Note that using per_rcpt with a non-readonly update mode in more than
2008 one ACL will cause the recipients to be double-counted. (The per_mail
2009 and per_byte options don't have this problem.)
2011 The handling of very low rates has changed slightly. If the computed rate
2012 is less than the event's count (usually one) then this event is the first
2013 after a long gap. In this case the rate is set to the same as this event's
2014 count, so that the first message of a spam run is counted properly.
2016 The major new feature is a mechanism for counting the rate of unique
2017 events. The new per_addr option counts the number of different
2018 recipients that someone has sent messages to in the last time period. It
2019 behaves like per_rcpt if all the recipient addresses are different, but
2020 duplicate recipient addresses do not increase the measured rate. Like
2021 the /count= option this is a general mechanism, so the per_addr option
2022 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2023 example, measure the rate that a client uses different sender addresses
2024 with the options per_mail/unique=$sender_address. There are further
2025 details in the main documentation.
2027 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2029 TF/06 Removed a few PCRE remnants.
2031 TF/07 Automatically extract Exim's version number from tags in the git
2032 repository when doing development or release builds.
2034 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2035 Bugzilla 879. Patch from Paul Fisher.
2037 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2038 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2041 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2043 PP/05 Variable $av_failed, true if the AV scanner deferred.
2044 Bugzilla 1078. Patch from John Horne.
2046 PP/06 Stop make process more reliably on build failure.
2047 Bugzilla 1087. Patch from Heiko Schlittermann.
2049 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2050 Bugzilla 1089. Patch from Heiko Schlittermann.
2052 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2053 Bugzilla 1131. Patch from Holger Weiß.
2055 PP/09 Handle IPv6 addresses with SPF.
2056 Bugzilla 860. Patch from Wolfgang Breyha.
2058 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2060 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2063 PP/11 match_* no longer expand right-hand-side by default.
2064 New compile-time build option, EXPAND_LISTMATCH_RHS.
2065 New expansion conditions, "inlist", "inlisti".
2067 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2069 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2071 PP/14 fix log_write() format string regression from TF/03.
2072 Bugzilla 1152. Patch from Dmitry Isaikin.
2078 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2080 PP/02 Harmonised TLS library version reporting; only show if debugging.
2081 Layout now matches that introduced for other libraries in 4.74 PP/03.
2083 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2085 PP/04 New "dns_use_edns0" global option.
2087 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2090 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2091 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2093 TK/01 Updated PolarSSL code to 0.14.2.
2094 Bugzilla 1097. Patch from Andreas Metzler.
2096 PP/07 Catch divide-by-zero in ${eval:...}.
2097 Fixes bugzilla 1102.
2099 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2102 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2103 format-string attack -- SECURITY: remote arbitrary code execution.
2105 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2106 time unintentionally subject to list matching rules, letting the header
2107 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2108 arbitrary string expansion). This allowed for information disclosure.
2110 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2111 INT_MIN/-1 -- value coerced to INT_MAX.
2117 NM/01 Workaround for PCRE version dependency in version reporting
2120 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2121 This fixes portability to compilers other than gcc, notably
2122 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2124 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2125 makefiles for portability to HP-UX and POSIX correctness.
2127 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2128 Also via indented variable definition in the Makefile.
2129 (Debugging by Oliver Heesakkers).
2131 PP/02 Restore caching of spamd results with expanded spamd_address.
2132 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2134 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2135 Improves build reliability. Fix from: Frank Elsner
2137 NM/02 Fix wide character breakage in the rfc2047 coding
2138 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2140 NM/03 Allow underscore in dnslist lookups
2141 Fixes bug 1026. Patch from Graeme Fowler
2143 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2144 Code patches from Adam Ciarcinski of NetBSD.
2146 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2149 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2150 is logged, to avoid truncation. Patch from John Horne.
2152 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2153 Patch from Jakob Hirsch.
2155 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2156 SQL string expansion failure details.
2157 Patch from Andrey Oktyabrski.
2159 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2160 Patch from Simon Arlott.
2162 PP/09 New lookups functionality failed to compile on old gcc which rejects
2163 extern declarations in function scope.
2164 Patch from Oliver Fleischmann
2166 PP/10 Use sig_atomic_t for flags set from signal handlers.
2167 Check getgroups() return and improve debugging.
2168 Fixed developed for diagnosis in bug 927 (which turned out to be
2171 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2172 Patch from Mark Zealey.
2174 PP/12 Bugzilla 1056: Improved spamd server selection.
2175 Patch from Mark Zealey.
2177 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2178 Based on patch from Heiko Schlittermann.
2180 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2181 Patch from Uwe Doering, sign-off by Michael Haardt.
2183 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2184 variable declaration deep within a block. Bug and patch from
2187 PP/15 lookups-Makefile IRIX compatibility coercion.
2189 PP/16 Make DISABLE_DKIM build knob functional.
2191 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2192 Patch by Simon Arlott
2194 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2195 variable argument macros. Our copy now differs from upstream.
2201 TF/01 Failure to get a lock on a hints database can have serious
2202 consequences so log it to the panic log.
2204 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2205 controlled using the smtp_confirmation log selector.
2207 TF/03 Include the error message when we fail to unlink a spool file.
2209 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2210 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2211 for maintaining out-of-tree patches for some time.
2213 PP/01 Bugzilla 139: Documentation and portability issues.
2214 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2215 Handle per-OS dynamic-module compilation flags.
2217 PP/02 Let /dev/null have normal permissions.
2218 The 4.73 fixes were a little too stringent and complained about the
2219 permissions on /dev/null. Exempt it from some checks.
2220 Reported by Andreas M. Kirchwitz.
2222 PP/03 Report version information for many libraries, including
2223 Exim version information for dynamically loaded libraries. Created
2224 version.h, now support a version extension string for distributors
2225 who patch heavily. Dynamic module ABI change.
2227 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2228 privilege escalation vulnerability whereby the Exim run-time user
2229 can cause root to append content of the attacker's choosing to
2232 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2235 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2236 If dropping privileges for untrusted macros, we disabled normal logging
2237 on the basis that it would fail; for the Exim run-time user, this is not
2238 the case, and it resulted in successful deliveries going unlogged.
2239 Fixed. Reported by Andreas Metzler.
2245 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2246 only prepend for the Resent-* case. Fixes regression introduced in
2247 Exim 4.70 by NM/22 for Bugzilla 607.
2249 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2250 increasing numbers of administrators be bitten by this.
2252 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2254 PP/03 Bugzilla 994: added openssl_options main configuration option.
2256 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2258 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2260 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2262 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2263 without a peer certificate, leading to a segfault because of an
2264 assumption that peers always have certificates. Be a little more
2265 paranoid. Problem reported by Martin Tscholak.
2267 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2268 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2269 NB: ClamAV planning to remove STREAM in "middle of 2010".
2270 CL also introduces -bmalware, various -d+acl logging additions and
2271 more caution in buffer sizes.
2273 PP/09 Implemented reverse_ip expansion operator.
2275 PP/10 Bugzilla 937: provide a "debug" ACL control.
2277 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2279 PP/12 Bugzilla 973: Implement --version.
2281 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2283 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2285 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2287 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2288 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2289 ignore trailing whitespace.
2291 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2293 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2296 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2297 Notification from Dr Andrew Aitchison.
2299 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2300 ExtendedDetectionInfo response format.
2301 Notification from John Horne.
2303 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2306 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2307 XSL and documented dependency on system catalogs, with examples of how
2310 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2313 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2314 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2315 configuration file which is writeable by the Exim user or group.
2317 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2318 of configuration files to cover files specified with the -C option if
2319 they are going to be used with root privileges, not just the default
2322 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2323 option (effectively making it always true).
2325 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2326 files to be used while preserving root privileges.
2328 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2329 that rogue child processes cannot use them.
2331 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2332 run-time user, instead of root.
2334 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2335 Exim run-time user without dropping privileges.
2337 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2338 result string, instead of calling string_vformat() twice with the same
2341 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2342 for other users. Others should always drop root privileges if they use
2343 -C on the command line, even for a whitelisted configure file.
2345 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2347 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2353 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2354 $data_path, and $header_path variables; fixed documentation bugs and
2357 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2358 exipick to access non-standard spools, including the "frozen" queue
2361 NM/01 Bugzilla 965: Support mysql stored procedures.
2362 Patch from Alain Williams
2364 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2366 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2367 Patch from Andreas Metzler
2369 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2370 Patch from Kirill Miazine
2372 NM/05 Bugzilla 671: Added umask to procmail example.
2374 JJ/03 installed exipick 20100323.0, fixing doc bug
2376 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2377 directory. Notification and patch from Dan Rosenberg.
2379 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2381 TK/02 Improve log output when DKIM signing operation fails.
2383 MH/01 Treat the transport option dkim_domain as a colon separated
2384 list, not as a single string, and sign the message with each element,
2385 omitting multiple occurences of the same signer.
2387 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2388 Bugzilla 985, 986. Patch by Simon Arlott
2390 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2391 Patch by Simon Arlott
2393 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2394 MBX locking. Notification from Dan Rosenberg.
2400 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2402 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2404 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2406 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2408 NM/04 Bugzilla 915: Flag broken perl installation during build.
2414 TK/01 Added patch by Johannes Berg that expands the main option
2415 "spamd_address" if it starts with a dollar sign.
2417 TK/02 Write list of recipients to X-Envelope-Sender header when building
2418 the mbox-format spool file for content scanning (suggested by Jakob
2421 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2422 (http://www.dcc-servers.net/) support via dccifd. Activated by
2423 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2425 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2426 by Mark Daniel Reidel <mr@df.eu>.
2428 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2429 When building exim an external PCRE library is now needed -
2430 PCRE is a system library on the majority of modern systems.
2431 See entry on PCRE_LIBS in EDITME file.
2433 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2434 conversation. Added nologin parameter to request.
2435 Patch contributed by Kirill Miazine.
2437 TF/01 Do not log submission mode rewrites if they do not change the address.
2439 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2441 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2442 log files in place. Contributed by Roberto Lima.
2444 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2446 TF/03 Bugzilla 615: When checking the local_parts router precondition
2447 after a local_part_suffix or local_part_prefix option, Exim now
2448 does not use the address's named list lookup cache, since this
2449 contains cached lookups for the whole local part.
2451 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2452 Robert Millan. Documentation is in experimental-spec.txt.
2454 TF/04 Bugzilla 668: Fix parallel build (make -j).
2456 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2458 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2459 Patch provided by Jan Srzednicki.
2461 TF/05 Leading white space used to be stripped from $spam_report which
2462 wrecked the formatting. Now it is preserved.
2464 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2465 that they are available at delivery time.
2467 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2469 TF/08 TLS error reporting now respects the incoming_interface and
2470 incoming_port log selectors.
2472 TF/09 Produce a more useful error message if an SMTP transport's hosts
2473 setting expands to an empty string.
2475 NM/06 Bugzilla 744: EXPN did not work under TLS.
2476 Patch provided by Phil Pennock.
2478 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2479 Patch provided by Richard Godbee.
2481 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2482 acl_smtp_notquit, added index entry.
2484 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2485 Patch provided by Eugene Bujak.
2487 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2488 accept(). Patch provided by Maxim Dounin.
2490 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2491 Patch provided by Phil Pennock.
2493 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2495 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2496 Patch provided by Brad "anomie" Jorsch.
2498 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2499 Patch provided by Dean Brooks.
2501 TK/05 Add native DKIM support (does not depend on external libraries).
2503 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2504 Patch provided by Graeme Fowler.
2506 NM/16 Bugzilla 851: Documentation example syntax fix.
2508 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2510 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2513 NM/19 Bugzilla 745: TLS version reporting.
2514 Patch provided by Phil Pennock.
2516 NM/20 Bugzilla 167: bool: condition support.
2517 Patch provided by Phil Pennock.
2519 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2520 clients. Patch provided by Phil Pennock.
2522 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2523 Patch provided by Brad "anomie" Jorsch.
2525 NM/23 Bugzilla 687: Fix misparses in eximstats.
2526 Patch provided by Heiko Schlittermann.
2528 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2529 Patch provided by Heiko Schlittermann.
2531 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2532 plus update to original patch.
2534 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2536 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2537 Patch provided by David Brownlee.
2539 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2541 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2543 NM/30 Bugzilla 888: TLS documentation bugfixes.
2545 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2547 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2548 Unlike the original bugzilla I have changed all shell scripts in src tree.
2550 NM/33 Bugzilla 898: Transport filter timeout fix.
2551 Patch by Todd Rinaldo.
2553 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2554 Patch by Serge Demonchaux.
2556 NM/35 Bugzilla 39: Base64 decode bug fixes.
2557 Patch by Jakob Hirsch.
2559 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2561 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2563 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2565 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2571 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2572 ALT-N's libdkim that I have put here:
2573 http://duncanthrax.net/exim-experimental/
2575 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2576 were called 'true' and it seems that C99 defines that as a reserved
2577 keyword to be used with 'bool' variable types. That means you could
2578 not include C99-style headers which use bools without triggering
2579 build errors in sieve.c.
2581 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2582 as mailq or other aliases. Changed the --help handling significantly
2583 to do whats expected. exim_usage() emits usage/help information.
2585 SC/01 Added the -bylocaldomain option to eximstats.
2587 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2589 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2591 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2597 PH/01 Another patch from the Sieve maintainer.
2599 PH/02 When an IPv6 address is converted to a string for single-key lookup
2600 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2601 dots are used instead of colons so that keys in lsearch files need not
2602 contain colons. This was done some time before quoting was made available
2603 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2604 (notated using the quote facility) so as to distinguish them from IPv4
2605 keys. This meant that lookups for IP addresses in host lists did not work
2606 for iplsearch lookups.
2608 This has been fixed by arranging for IPv6 addresses to be expressed with
2609 colons if the lookup type is iplsearch. This is not incompatible, because
2610 previously such lookups could never work.
2612 The situation is now rather anomalous, since one *can* have colons in
2613 ordinary lsearch keys. However, making the change in all cases is
2614 incompatible and would probably break a number of configurations.
2616 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2619 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2620 conversion specification without a maximum field width, thereby enabling
2621 a rogue spamd server to cause a buffer overflow. While nobody in their
2622 right mind would setup Exim to query an untrusted spamd server, an
2623 attacker that gains access to a server running spamd could potentially
2624 exploit this vulnerability to run arbitrary code as the Exim user.
2626 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2627 $primary_hostname instead of what libspf2 thinks the hosts name is.
2629 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2630 a directory entry by the name of the lookup key. Previously, if a
2631 symlink pointed to a non-existing file or a file in a directory that
2632 Exim lacked permissions to read, a lookup for a key matching that
2633 symlink would fail. Now it is enough that a matching directory entry
2634 exists, symlink or not. (Bugzilla 503.)
2636 PH/03 The body_linecount and body_zerocount variables are now exported in the
2639 PH/04 Added the $dnslist_matched variable.
2641 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2642 This means they are set thereafter only if the connection becomes
2645 PH/06 Added the client_condition to authenticators so that some can be skipped
2646 by clients under certain conditions.
2648 PH/07 The error message for a badly-placed control=no_multiline_responses left
2649 "_responses" off the end of the name.
2651 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2653 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2654 (without spaces) instead of just copying the configuration text.
2656 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2658 PH/11 Added $max_received_linelength.
2660 PH/12 Added +ignore_defer and +include_defer to host lists.
2662 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2663 way in which PCRE > 7.0 is built.
2665 PH/14 Implemented queue_only_load_latch.
2667 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2668 MAIL command. The effect was to mangle the value on 64-bit systems.
2670 PH/16 Another patch from the Sieve maintainer.
2672 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2674 PH/18 If a system quota error occurred while trying to create the file for
2675 a maildir delivery, the message "Mailbox is full" was not appended to the
2676 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2677 only to a quota excession during the actual writing of the file.
2679 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2680 characters?) which causes problems in log lines. The DN values are now
2681 passed through string_printing() before being added to log lines.
2683 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2684 and InterBase are left for another time.)
2686 PH/21 Added message_body_newlines option.
2688 PH/22 Guard against possible overflow in moan_check_errorcopy().
2690 PH/23 POSIX allows open() to be a macro; guard against that.
2692 PH/24 If the recipient of an error message contained an @ in the local part
2693 (suitably quoted, of course), incorrect values were put in $domain and
2694 $local_part during the evaluation of errors_copy.
2700 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2701 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2704 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2705 issue a MAIL command.
2707 PH/02 In an ACL statement such as
2709 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2711 if a client was not listed at all, or was listed with a value other than
2712 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2713 the condition was not true (as it should be), so access was not denied.
2714 The bug was that the ! inversion was incorrectly passed on to the second
2715 item. This has been fixed.
2717 PH/03 Added additional dnslists conditions == and =& which are different from
2718 = and & when the dns lookup returns more than one IP address.
2720 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2721 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2723 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2724 FSYNC, which compiles an option called disable_fsync that allows for
2725 bypassing fsync(). The documentation is heavily laced with warnings.
2727 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2729 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2730 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2731 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2732 including adding "make clean"; (3) Added -fPIC when compiling the test
2733 dynamically loaded module, to get rid of a warning.
2735 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2736 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2737 = 0s. The bug is otherwise harmless.
2739 PH/07 There was a bug in the dovecot authenticator such that the value of
2740 $auth1 could be overwritten, and so not correctly preserved, after a
2741 successful authentication. This usually meant that the value preserved by
2742 the server_setid option was incorrect.
2744 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2746 PH/09 Installed PCRE release 7.0.
2748 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2749 run for batched SMTP input. It is now run at the start of every message
2750 in the batch. While fixing this I discovered that the process information
2751 (output by running exiwhat) was not always getting set for -bs and -bS
2752 input. This is fixed, and it now also says "batched" for BSMTP.
2754 PH/11 Added control=no_pipelining.
2756 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2757 patch, slightly modified), and move the expansion of helo_data till after
2758 the connection is made in the smtp transport (so it can use these
2761 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2763 PH/14 Added log_selector = +pid.
2765 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2767 PH/16 Add ${if forany and ${if forall.
2769 PH/17 Added dsn_from option to vary the From: line in DSNs.
2771 PH/18 Flush SMTP output before performing a callout, unless control =
2772 no_callout_flush is set.
2774 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2775 was true (the default) a successful delivery failed to delete the retry
2776 item, thus causing premature timeout of the address. The bug is now
2779 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2781 PH/21 Long custom messages for fakedefer and fakereject are now split up
2782 into multiline responses in the same way that messages for "deny" and
2783 other ACL rejections are.
2785 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2786 with slight modification.
2788 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2789 draft, changing the syntax and factoring some duplicate code".
2791 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2792 for deliveries of the second and subsequent messages over the same SMTP
2795 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2796 ${reduce, with only minor "tidies".
2798 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2800 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2801 expansion side effects.
2803 PH/27 When a message times out after an over-quota error from an Exim-imposed
2804 quota, the bounce message says "mailbox is full". This message was not
2805 being given when it was a system quota that was exceeded. It now should
2808 MH/03 Made $recipients available in local_scan(). local_scan() already has
2809 better access to the recipient list through recipients_list[], but
2810 $recipients can be useful in postmaster-provided expansion strings.
2812 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2813 in the case of a MAIL command with additional options following the
2814 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2815 were accidentally chopped off.
2817 PH/29 SMTP synchronization checks are implemented when a command is read -
2818 there is a check that no more input is waiting when there shouldn't be
2819 any. However, for some commands, a delay in an ACL can mean that it is
2820 some time before the response is written. In this time, more input might
2821 arrive, invalidly. So now there are extra checks after an ACL has run for
2822 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2823 pipelining has not been advertised.
2825 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2827 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2828 returned in the SMTP response when smtp_return_error_details was false.
2829 This has been fixed.
2831 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2832 instead of the C I/O that was originally supplied, because problems were
2833 reported on Solaris.
2835 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2836 Exim which did not show up earlier: it was assuming that a call to
2837 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2838 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2839 was a macro that became an assignment, so it seemed to work. This has
2840 changed to a proper function call with a void return, hence the compile
2841 error. Exim's code has been fixed.
2843 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2846 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2848 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2850 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2851 because it turns out to be too big or there is a timeout) there is no
2852 "Completed" line in the log. When some messages of this type were
2853 selected by exigrep, they were listed as "not completed". Others were
2854 picked up by some special patterns. I have improved the selection
2855 criteria to be more general.
2857 PH/38 The host_find_failed option in the manualroute router can now be set
2858 to "ignore", to completely ignore a host whose IP address cannot be
2859 found. If all hosts are ignored, the behaviour is controlled by the new
2860 host_all_ignored option.
2862 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2863 homing or because of multiple MX records with /mx) generated more than
2864 one IP address, and the following item turned out to be the local host,
2865 all the secondary addresses of the first item were incorrectly removed
2866 from the list, along with the local host and any following hosts (which
2867 is what is supposed to happen).
2869 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2870 whoever called Exim into the -H file. In the case of the daemon it was
2871 behaving confusingly. When first started, it used values for whoever
2872 started the daemon, but after a SIGHUP it used the Exim user (because it
2873 calls itself on a restart). I have changed the code so that it now always
2876 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2877 message are rejected with the same error (e.g. no authentication or bad
2878 sender address), and a DATA command is nevertheless sent (as can happen
2879 with PIPELINING or a stupid MUA), the error message that was given to the
2880 RCPT commands is included in the rejection of the DATA command. This is
2881 intended to be helpful for MUAs that show only the final error to their
2884 PH/42 Another patch from the Sieve maintainer.
2886 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2887 Eximstats - Fixed some broken HTML links and added missing column headers
2889 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2890 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2892 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2894 PH/43 Yet another patch from the Sieve maintainer.
2896 PH/44 I found a way to check for a TCP/IP connection going away before sending
2897 the response to the final '.' that terminates a message, but only in the
2898 case where the client has not sent further data following the '.'
2899 (unfortunately, this is allowed). However, in many cases there won't be
2900 any further data because there won't be any more messages to send. A call
2901 to select() can be used: if it shows that the input is "ready", there is
2902 either input waiting, or the socket has been closed. An attempt to read
2903 the next input character can distinguish the two cases. Previously, Exim
2904 would have sent an OK response which the client would never have see.
2905 This could lead to message repetition. This fix should cure that, at
2906 least in a lot of common cases.
2908 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2909 advertised in response to EHLO.
2915 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2916 fixed by 4.65/MH/01 (is this a record?) are fixed:
2918 (i) An empty string was always treated as zero by the numeric comparison
2919 operators. This behaviour has been restored.
2921 (ii) It is documented that the numeric comparison operators always treat
2922 their arguments as decimal numbers. This was broken in that numbers
2923 starting with 0 were being interpreted as octal.
2925 While fixing these problems I realized that there was another issue that
2926 hadn't been noticed. Values of message_size_limit (both the global option
2927 and the transport option) were treated as octal if they started with 0.
2928 The documentation was vague. These values are now always treated as
2929 decimal, and I will make that clear in the documentation.
2935 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2936 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2939 MH/01 Don't check that the operands of numeric comparison operators are
2940 integers when their expansion is in "skipping" mode (fixes bug
2941 introduced by 4.64-PH/07).
2943 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2944 child addresses, Exim now panics and dies. Previously, because the count
2945 is held in a short int, deliveries were likely to be lost. As such a
2946 large number of recipients for a single message is ridiculous
2947 (performance will be very, very poor), I have chosen to impose a limit
2948 rather than extend the field.
2954 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2955 leftover -K file (the existence of which was triggered by #402).
2956 While we were at it, introduced process PID as part of the -K
2957 filename. This should rule out race conditions when creating
2960 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2961 processing considerably. Previous code took too long for large mails,
2962 triggering a timeout which in turn triggers #401.
2964 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2965 in the DK code in transports.c. sendfile() is not really portable,
2966 hence the _LINUX specificness.
2968 TF/01 In the add_headers option to the mail command in an Exim filter,
2969 there was a bug that Exim would claim a syntax error in any
2970 header after the first one which had an odd number of characters
2973 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2974 callout verification, Exim cached a "reject" for the entire domain. This
2975 is correct for most verifications, but it is not correct for a recipient
2976 verification with use_sender or use_postmaster set, because in that case
2977 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2978 case of MAIL FROM:<> rejection from other early rejections (e.g.
2979 rejection of HELO). When verifying a recipient using a non-null MAIL
2980 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2981 Whatever the result of the callout, the value of the domain cache is
2982 left unchanged (for any other kind of callout, getting as far as trying
2983 RCPT means that the domain itself is ok).
2985 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2988 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2989 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2990 session, the daemon ignores ECONNECT errors and logs others; it now
2991 ignores EPIPE as well.
2993 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2994 (quoted-printable decoding).
2996 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2997 later the small subsequent patch to fix an introduced bug.
2999 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3001 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3003 PH/08 An error is now given if message_size_limit is specified negative.
3005 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3006 to be given (somewhat) arbitrary names.
3008 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3011 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3012 miscellaneous code fixes
3014 PH/10 Added the log_reject_target ACL modifier to specify where to log
3017 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3018 hostname. This is wrong, because it relates to the incoming message (and
3019 probably the interface on which it is arriving) and not to the outgoing
3020 callout (which could be using a different interface). This has been
3021 changed to use the value of the helo_data option from the smtp transport
3022 instead - this is what is used when a message is actually being sent. If
3023 there is no remote transport (possible with a router that sets up host
3024 addresses), $smtp_active_hostname is used.
3026 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3027 tweaks were necessary in order to get it to work (see also 21 below):
3028 (a) The code assumed that strncpy() returns a negative number on buffer
3029 overflow, which isn't the case. Replaced with Exim's string_format()
3031 (b) There were several signed/unsigned issues. I just did the minimum
3032 hacking in of casts. There is scope for a larger refactoring.
3033 (c) The code used strcasecmp() which is not a standard C function.
3034 Replaced with Exim's strcmpic() function.
3035 (d) The code set only $1; it now sets $auth1 as well.
3036 (e) A simple test gave the error "authentication client didn't specify
3037 service in request". It would seem that Dovecot has changed its
3038 interface. Fortunately there's a specification; I followed it and
3039 changed what the client sends and it appears to be working now.
3041 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3044 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3045 address A is aliased to B and C, where B exists and C does not. Without
3046 -v the output is "A verified" because verification stops after a
3047 successful redirection if more than one address is generated. However,
3048 with -v the child addresses are also verified. Exim was outputting "A
3049 failed to verify" and then showing the successful verification for C,
3050 with its parentage. It now outputs "B failed to verify", showing B's
3051 parentage before showing the successful verification of C.
3053 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3054 look up a TXT record in a specific list after matching in a combined
3057 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3058 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3059 they consult the DNS. I had assumed they would set it the way they
3060 wanted; and indeed my experiments on Linux seem to show that in some
3061 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3062 To be on the safe side, however, I have now made the interface to
3063 host_find_byname() similar to host_find_bydns(), with an argument
3064 containing the DNS resolver options. The host_find_byname() function now
3065 sets these options at its start, just as host_find_bydns() does. The smtp
3066 transport options dns_qualify_single and dns_search_parents are passed to
3067 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3068 of host_find_byname() use the default settings of RES_DEFNAMES
3069 (qualify_single) but not RES_DNSRCH (search_parents).
3071 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3072 spool_read_header() do less string testing, by means of a preliminary
3073 switch on the second character of optional "-foo" lines. (This is
3074 overdue, caused by the large number of possibilities that now exist.
3075 Originally there were few.) While I was there, I also converted the
3076 str(n)cmp tests so they don't re-test the leading "-" and the first
3077 character, in the hope this might squeeze out yet more improvement.
3079 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3080 flag allowing group syntax was set by the header_syntax check but not
3081 turned off, possible causing trouble later; (2) The flag was not being
3082 set at all for the header_verify test, causing "group"-style headers to
3083 be rejected. I have now set it in this case, and also caused header_
3084 verify to ignore an empty address taken from a group. While doing this, I
3085 came across some other cases where the code for allowing group syntax
3086 while scanning a header line wasn't quite right (mostly, not resetting
3087 the flag correctly in the right place). These bugs could have caused
3088 trouble for malformed header lines. I hope it is now all correct.
3090 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3091 with the "reply" argument non-NULL. The code, however (which originally
3092 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3093 but it didn't always do it. This confused somebody who was copying the
3094 code for some other use. I have removed all the tests.
3096 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3097 feature that was used to support insecure browsers during the U.S. crypto
3098 embargo. It requires special client support, and Exim is probably the
3099 only MTA that supported it -- and would never use it because real RSA is
3100 always available. This code has been removed, because it had the bad
3101 effect of slowing Exim down by computing (never used) parameters for the
3102 RSA_EXPORT functionality.
3104 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3105 authenticator to fail if there's a tab character in the incoming data
3106 (there should never be unless someone is messing about, as it's supposed
3107 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3108 if the connection is using TLS or if the remote IP is the same as the
3109 local IP, and the "valid-client-cert option" if a client certificate has
3112 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3113 authenticators. This can be used for authorization after authentication
3114 succeeds. (In the case of plaintext, it servers for both authentication
3117 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3118 if any retry times were supplied.
3120 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3121 connection, where there is no client IP address to check. In this
3122 situation, the verify now always succeeds.
3124 PH/25 Applied John Jetmore's -Mset patch.
3126 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3128 PH/27 In a string expansion for a processed (not raw) header when multiple
3129 headers of the same name were present, leading whitespace was being
3130 removed from all of them, but trailing whitespace was being removed only
3131 from the last one. Now trailing whitespace is removed from each header
3132 before concatenation. Completely empty headers in a concatenation (as
3133 before) are ignored.
3135 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3136 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3138 PH/29 [Removed. This was a change that I later backed out, and forgot to
3139 correct the ChangeLog entry (that I had efficiently created) before
3140 committing the later change.]
3142 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3143 address errors (4xx response to RCPT) over the same connection as other
3144 messages routed to the same hosts. Such deliveries are always "forced",
3145 so retry times are not inspected. This resulted in far too many retries
3146 for the affected addresses. The effect occurred only when there were more
3147 hosts than the hosts_max_try setting in the smtp transport when it had
3148 the 4xx errors. Those hosts that it had tried were not added to the list
3149 of hosts for which the message was waiting, so if all were tried, there
3150 was no problem. Two fixes have been applied:
3152 (i) If there are any address or message errors in an SMTP delivery, none
3153 of the hosts (tried or untried) are now added to the list of hosts
3154 for which the message is waiting, so the message should not be a
3155 candidate for sending over the same connection that was used for a
3156 successful delivery of some other message. This seems entirely
3157 reasonable: after all the message is NOT "waiting for some host".
3158 This is so "obvious" that I'm not sure why it wasn't done
3159 previously. Hope I haven't missed anything, but it can't do any
3160 harm, as the worst effect is to miss an optimization.
3162 (ii) If, despite (i), such a delivery is accidentally attempted, the
3163 routing retry time is respected, so at least it doesn't keep
3164 hammering the server.
3166 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3167 in ${readsocket because some servers need this prod.
3169 PH/32 Added some extra debug output when updating a wait-xxx database.
3171 PH/33 The hint "could be header name not terminated by colon", which has been
3172 given for certain expansion errors for a long time, was not being given
3173 for the ${if def:h_colon_omitted{... case.
3175 PH/34 The spec says: "With one important exception, whenever a domain list is
3176 being scanned, $domain contains the subject domain." There was at least
3177 one case where this was not true.
3179 PH/35 The error "getsockname() failed: connection reset by peer" was being
3180 written to the panic log as well as the main log, but it isn't really
3181 panic-worthy as it just means the connection died rather early on. I have
3182 removed the panic log writing for the ECONNRESET error when getsockname()
3185 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3186 runs only) independently of the message's sender address. This meant
3187 that, if the 4xx error was in fact related to the sender, a different
3188 message to the same recipient with a different sender could confuse
3189 things. In particular, this can happen when sending to a greylisting
3190 server, but other circumstances could also provoke similar problems.
3191 I have changed the default so that the retry time for these errors is now
3192 based a combination of the sender and recipient addresses. This change
3193 can be overridden by setting address_retry_include_sender=false in the
3196 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3197 remote server are returned as part of bounce messages. This was not
3198 happening for LMTP over a pipe (the lmtp transport), but now it is the
3199 same for both kinds of LMTP.
3201 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3202 in header lines that were in fact CNAMEs. This is no longer the case.
3204 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3205 and queue runs started by the daemon processed all messages. This has
3206 been fixed so that -R and -S can now usefully be given with -q<time>.
3208 PH/40 Import PCRE release 6.7 (fixes some bugs).
3210 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3212 PH/42 Give an error if -q is specified more than once.
3214 PH/43 Renamed the variables $interface_address and $interface_port as
3215 $received_ip_address and $received_port, to make it clear that these
3216 values apply to message reception, and not to the outgoing interface when
3217 a message is delivered. (The old names remain recognized, of course.)
3219 PH/44 There was no timeout on the connect() call when using a Unix domain
3220 socket in the ${readsocket expansion. There now is.
3222 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3223 be meaningful with "accept".
3225 SC/01 Eximstats V1.43
3226 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3228 SC/02 Eximstats V1.44
3229 Use a glob alias rather than an array ref in the generated
3230 parser. This improves both readability and performance.
3232 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3233 Collect SpamAssassin and rejection statistics.
3234 Don't display local sender or destination tables unless
3235 there is data to show.
3236 Added average volumes into the top table text output.
3238 SC/04 Eximstats V1.46
3239 Collect data on the number of addresses (recipients)
3240 as well as the number of messages.
3242 SC/05 Eximstats V1.47
3243 Added 'Message too big' to the list of mail rejection
3244 reasons (thanks to Marco Gaiarin).
3246 SC/06 Eximstats V1.48
3247 Mainlog lines which have GMT offsets and are too short to
3248 have a flag are now skipped.
3250 SC/07 Eximstats V1.49 (Alain Williams)
3251 Added the -emptyok flag.
3253 SC/08 Eximstats V1.50
3254 Fixes for obtaining the IP address from reject messages.
3256 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3257 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3258 whitespace changes from 4.64-PH/27
3260 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3263 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3264 are found, allow negative numbers in numeric criteria)
3266 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3268 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3269 to match changes made in 4.64-PH/43
3271 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3273 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3274 "domain" is actually an IP address, and force a failure. This locks out
3275 those revolvers/nameservers that support "A-for-A" lookups, in
3276 contravention of the specifications.
3278 PH/48 When a host name was looked up from an IP address, and the subsequent
3279 forward lookup of the name timed out, the host name was left in
3280 $sender_host_name, contrary to the specification.
3282 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3283 restricted to single-key lookups, Exim was not diagnosing an error if
3284 * or *@ was used with a query-style lookup.
3286 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3288 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3289 long ago, but noone interested enough thought of it. Let's just say that
3290 the "1.1" means that there are some new functions that weren't there at
3291 some point in the past.
3293 PH/51 Error processing for expansion failure of helo_data from an smtp
3294 transport during callout processing was broken.
3296 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3297 tested/used via the -bh/-bhc/-bs options.
3299 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3300 bug, fixed in subsequent PCRE releases).
3302 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3303 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3305 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3311 SC/01 Use a glob alias rather than an array ref in eximstats generated
3312 parser. This improves both readability and performance.
3314 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3315 Don't display local sender or destination tables in eximstats unless
3316 there is data to show.
3317 Added average volumes into the eximstats top table text output.
3319 SC/03 Collect data on the number of addresses (recipients) as well
3320 as the number of messages in eximstats.
3322 TF/01 Correct an error in the documentation for the redirect router. Exim
3323 does (usually) call initgroups() when daemonizing.
3325 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3326 with consistent privilege compared to when running as a daemon.
3328 TF/03 Note in the spec that $authenticated_id is not set for local
3329 submissions from trusted users.
3331 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3332 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3334 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3335 by adding some example configuration directives to the default
3336 configuration file. A little bit of work is required to uncomment the
3337 directives and define how usernames and passwords are checked, but
3338 there is now a framework to start from.
3340 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3341 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3342 without this. I don't know how relevant this is to other LDAP libraries.
3344 PH/02 Add the verb name to the "unknown ACL verb" error.
3346 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3348 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3350 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3351 directory not expanded when it should be if an expanded home directory
3352 was set for the address (which is overridden by the transport).
3354 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3357 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3358 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3359 because it is too late at that time, and has no effect.
3361 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3362 security issue with \' (bugzilla #107). I could not use the
3363 PQescapeStringConn() function, because it needs a PGconn value as one of
3366 PH/08 When testing addresses using -bt, indicate those final addresses that
3367 are duplicates that would not cause an additional delivery. At least one
3368 person was confused, thinking that -bt output corresponded to deliveries.
3369 (Suppressing duplicates isn't a good idea as you lose the information
3370 about possibly different redirections that led to the duplicates.)
3372 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3373 systems where poll() doesn't work, in particular OS X.
3375 PH/10 Added more information to debugging output for retry time not reached.
3377 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3378 operations in malware.c.
3380 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3383 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3384 syslog_duplication set false, log lines that would normally be written
3385 both the the main log and to the reject log were not written to syslog at
3388 PH/14 In the default configuration, change the use of "message" in ACL warn
3389 statements to "add_header".
3391 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3392 not followed by a command (e.g. "seen endif").
3394 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3395 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3398 PH/17 Added extra conditions to the default value of delay_warning_condition
3402 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3403 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3404 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3407 The Auto-Submitted: and various List- headers are standardised, whereas I
3408 don't think Precedence: ever was.
3410 PH/18 Refactored debugging code in route_finduser() to show more information,
3411 in particular, the error code if getpwnam() issues one.
3413 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3414 This is apparently needed in addition to the PH/07 change above to avoid
3415 any possible encoding problems.
3417 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3418 but not after initializing Perl.
3420 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3421 output them only if debugging. By default they are written stderr,
3422 apparently, which is not desirable.
3424 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3427 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3430 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3432 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3433 authenticated or an ident call has been made. Suppress the default
3434 values for $authenticated_id and $authenticated_sender (but permit -oMai
3435 and -oMas) when testing with -bh.
3437 PH/24 Re-jigged the order of the tests in the default configuration so that the
3438 tests for valid domains and recipients precede the DNS black list and CSA
3439 tests, on the grounds that those ones are more expensive.
3441 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3442 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3443 This bug exists in every version of Exim that I still have, right back to
3446 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3447 However, an attempt to turn on case-sensitivity in a regex key by
3448 including (?-i) didn't work because the subject string was already
3449 lowercased, and the effects were non-intuitive. It turns out that a
3450 one-line patch can be used to allow (?-i) to work as expected.
3456 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3457 other effects) broke the use of negated acl sub-conditions.
3459 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3462 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3463 "Deny" causes Exim to reject the incoming connection with a 554 error.
3464 Unfortunately, if there is a major crisis, such as a disk failure,
3465 tcp-wrappers gives "deny", whereas what one would like would be some
3466 kind of temporary error. A kludge has been added to help with this.
3467 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3468 554 error is used if errno is still zero or contains ENOENT (which occurs
3469 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3472 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3474 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3475 errors. Otherwise a message that provokes a temporary error (when other
3476 messages do not) can cause a whole host to time out.
3478 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3479 addresses were routed directly to files or pipes from a redirect router.
3480 File deliveries just didn't batch; pipe deliveries might have suffered
3483 PH/06 A failure to get a lock for a hints database would erroneously always say
3484 "Failed to get write lock", even when it was really a read lock.
3486 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3487 of 0600. This has been changed to use the value of the lockfile_mode
3488 option (which defaults to 0600).
3490 PH/08 Applied small patch from the Sieve maintainer.
3492 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3493 folder from quota calculations, a direct delivery into this folder messed
3494 up the contents of the maildirsize file. This was because the regex was
3495 used only to exclude .Trash (or whatever) when the size of the mailbox
3496 was calculated. There was no check that a delivery was happening into an
3497 excluded directory. This bug has been fixed by ignoring all quota
3498 processing for deliveries into excluded directories.
3500 PH/10 Added the maildirfolder_create_regex option to appendfile.
3506 PH/01 The code for finding all the local interface addresses on a FreeBSD
3507 system running IPv6 was broken. This may well have applied to all BSD
3508 systems, as well as to others that have similar system calls. The broken
3509 code found IPv4 interfaces correctly, but gave incorrect values for the
3510 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3511 that it would not match correctly against @[] and not recognize the IPv6
3514 PH/02 The ipliteral router was not recognizing addresses of the form user@
3515 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3517 PH/03 Added disable_ipv6.
3519 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3520 decoded form, because it is most often used to construct To: headers
3521 lines in autoreplies, and the decoded form may well be syntactically
3522 invalid. However, $reply_address has leading white space removed, and all
3523 newlines turned into spaces so that the autoreply transport does not
3526 PH/05 If group was specified without a user on a router, and no group or user
3527 was specified on a transport, the group from the router was ignored.
3529 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3530 for visible compile-time settings that can be used to change these
3531 numbers, for those that want even more. Backwards compatibility with old
3532 spool files has been maintained. However, going back to a previous Exim
3533 release will lost any variables that are in spool files.
3535 PH/07 Two small changes when running in the test harness: increase delay when
3536 passing a TCP/IP connection to a new process, in case the original
3537 process has to generate a bounce, and remove special handling of
3538 127.0.0.2 (sic), which is no longer necessary.
3540 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3541 be the same on different OS.
3543 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3546 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3547 whether --show-vars was specified or not
3549 JJ/02 exipick: Added support for new ACL variable spool format introduced
3552 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3553 syntactically invalid From: or Reply-to: line, and a filter used this to
3554 generate an autoreply, and therefore failed to obtain an address for the
3555 autoreply, Exim could try to deliver to a non-existent relative file
3556 name, causing unrelated and misleading errors. What now happens is that
3557 it logs this as a hard delivery error, but does not attempt to create a
3560 PH/11 The exinext utility has a -C option for testing purposes, but although
3561 the given file was scanned by exinext itself; it wasn't being passed on
3562 when Exim was called.
3564 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3565 an end-of-file indication when reading a command response.
3567 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3568 compiled. In many other places in Exim, IPv6 addresses are always
3569 recognized, so I have changed this. It also means that IPv4 domain
3570 literals of the form [IPV4:n.n.n.n] are now always recognized.
3572 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3573 used if the router is not running as root, for example, when verifying at
3574 ACL time, or when using -bh. The debugging output from this situation was
3575 non-existent - all you got was a failure to exec. I have made two
3578 (a) Failures to set uid/gid, the current directory, or a process leader
3579 in a subprocess such as that created by queryprogram now generate
3580 suitable debugging output when -d is set.
3582 (b) The queryprogram router detects when it is not running as root,
3583 outputs suitable debugging information if -d is set, and then runs
3584 the subprocess without attempting to change uid/gid.
3586 PH/15 Minor change to Makefile for building test_host (undocumented testing
3589 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3590 additional section of a DNS packet that returns MX or SRV records.
3591 Instead, it always explicitly searches for A/AAAA records. This avoids
3592 major problems that occur when a DNS server includes only records of one
3593 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3594 fixed another bug: if SRV records were looked up and the corresponding
3595 address records were *not* found in the additional section, the port
3596 values from the SRV records were lost.
3598 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3599 using the correct key (the original address) when searching the retry
3600 rules in order to find which one to use for generating the retry hint.
3602 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3603 adding the default one. Similarly, if it contains a Reply-To: header, the
3604 errors_reply_to option, if set, is not used.
3606 PH/19 When calculating a retry time, Exim used to measure the "time since
3607 failure" by looking at the "first failed" field in the retry record. Now
3608 it does not use this if it is later than than the arrival time of the
3609 message. Instead it uses the arrival time. This makes for better
3610 behaviour in cases where some deliveries succeed, thus re-setting the
3611 "first failed" field. An example is a quota failure for a huge message
3612 when small messages continue to be delivered. Without this change, the
3613 "time since failure" will always be short, possible causing more frequent
3614 delivery attempts for the huge message than are intended.
3615 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3617 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3618 $1, $2, $3) because the numerical variables can be reset during some
3619 expansion items (e.g. "match"), thereby losing the authentication data.
3621 PH/21 Make -bV show the size of off_t variables so that the test suite can
3622 decide whether to run tests for quotas > 2G.
3624 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3625 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3626 filecount value is greater than 2G or if a quota value is greater than 2G
3627 on a system where the size of off_t is not greater than 4, a panic error
3630 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3631 never match. The debug and -bh output now contains an explicit error
3632 message indicating a malformed IPv4 address or mask.
3634 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3635 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3636 PH/23 above applies.
3638 PH/25 Do not write to syslog when running in the test harness. The only
3639 occasion when this arises is a failure to open the main or panic logs
3640 (for which there is an explicit test).
3642 PH/26 Added the /no_tell option to "control=freeze".
3644 PH/27 If a host name lookup failed very early in a connection, for example, if
3645 the IP address matched host_lookup and the reverse lookup yielded a name
3646 that did not have a forward lookup, an error message of the form "no IP
3647 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3648 could be logged. Now it outputs the IP address instead of "NULL".
3650 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3651 allows the sender and the authenticated sender to be set when
3652 submitting a message from within Exim. Since child_open_exim() is
3653 documented for local_scan(), the new function should be too.
3655 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3656 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3657 results in an empty string is now treated as unset.
3659 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3661 PH/31 Added sender_verify_fail logging option.
3663 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3664 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3665 tidied the source and removed it altogether.
3667 PH/33 When a queue run was abandoned because the load average was too high, a
3668 log line was always written; now it is written only if the queue_run log
3669 selector is set. In addition, the log line for abandonment now contains
3670 information about the queue run such as the pid. This is always present
3671 in "start" and "stop" lines but was omitted from the "abandon" line.
3673 PH/34 Omit spaces between a header name and the colon in the error message that
3674 is given when verify = headers_syntax fails (if there are lots of them,
3675 the message gets confusing).
3677 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3678 names, as there are now some PTR records that contain slashes. This check
3679 is only to protect against broken name servers that fall over on strange
3680 characters, so the fact that it applies to all lookups doesn't matter.
3682 PH/36 Now that the new test suite is complete, we can remove some of the
3683 special code in Exim that was needed for the old test suite. For example,
3684 sorting DNS records because real resolvers return them in an arbitrary
3685 order. The new test suite's fake resolver always returns records in the
3688 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3689 bounces) except when queue_only is set, to avoid logging races between
3690 the different processes.
3692 PH/38 Panic-die if .include specifies a non-absolute path.
3694 PH/39 A tweak to the "H" retry rule from its user.
3696 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3697 a label. They prevented compilation on older perls.
3699 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3700 a warning to be raised on newish perls.
3702 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3703 on queue. Changes to match documented behaviour of showing count of
3704 messages matching specified criteria.
3706 PH/40 Changed the default ident timeout from 30s to 5s.
3708 PH/41 Added support for the use of login_cap features, on those BSD systems
3709 that have them, for controlling the resources used by pipe deliveries.
3711 PH/42 The content-scanning code uses fopen() to create files in which to put
3712 message data. Previously it was not paying any attention to the mode of
3713 the files. Exim runs with umask(0) because the rest of the code creates
3714 files with open(), and sets the required mode explicitly. Thus, these
3715 files were ending up world-writeable. This was not a big issue, because,
3716 being within the spool directory, they were not world-accessible. I have
3717 created a function called modefopen, which takes an additional mode
3718 argument. It sets umask(777), creates the file, chmods it to the required
3719 mode, then resets the umask. All the relevant calls to fopen() in the
3720 content scanning code have been changed to use this function.
3722 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3723 to 24 hours. This avoids potential overflow problems when processing G
3724 and H retry rules. I suspect nobody ever tinkers with this value.
3726 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3728 PH/45 When the plaintext authenticator is running as a client, the server's
3729 challenges are checked to ensure they are valid base64 strings. By
3730 default, the authentication attempt is cancelled if an invalid string is
3731 received. Setting client_ignore_invalid_base64 true ignores these errors.
3732 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3733 they are received. Thus, the responses can be made to depend on the
3734 challenges. If an invalid string is ignored, an empty string is placed in
3737 PH/46 Messages that are created by the autoreply transport now contains a
3738 References: header, in accordance with RFCs 2822 and 3834.
3740 PH/47 Added authenticated_sender_force to the smtp transport.
3742 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3744 PH/49 Installed latest patch from the Sieve maintainer.
3746 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3747 was also set, the appendfile transport was unnecessarily scanning a
3748 directory of message files (e.g. for maildir delivery) to find the count
3749 of files (along with the size), even though it did not need this
3750 information. It now does the scan only if it needs to find either the
3751 size of the count of files.
3753 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3755 PH/52 Two bugs concerned with error handling when the smtp transport is
3758 (i) Exim was not creating retry information for temporary errors given
3759 for individual recipients after the DATA command when the smtp transport
3760 was used in LMTP mode. This meant that they could be retried too
3761 frequently, and not timed out correctly.
3763 (ii) Exim was setting the flag that allows error details to be returned
3764 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3765 recipients that were returned after the DATA command.
3767 PH/53 This is related to PH/52, but is more general: for any failing address,
3768 when detailed error information was permitted to be returned to the
3769 sender, but the error was temporary, then after the final timeout, only
3770 "retry timeout exceeded" was returned. Now it returns the full error as
3771 well as "retry timeout exceeded".
3773 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3774 do this, and (what is worse) MTAs that accept it.
3776 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3777 will now be deprecated.
3779 PH/56 New os.c-cygwin from the Cygwin maintainer.
3781 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3782 formats (previously only available in exim formats via -bpr, -bpru,
3783 and -bpra. Now also available in native and exiqgrep formats)
3785 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3786 with very large, slow to parse queues
3788 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3790 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3792 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3793 responses to authentication challenges, though it was showing the
3794 challenges; (ii) I've removed the CR characters from the debug output for
3797 PH/58 Allow for the insertion of a newline as well as a space when a string
3798 is turned into more than one encoded-word during RFC 2047 encoding. The
3799 Sieve code now uses this.
3801 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3802 data_4xx, lost_connection, tls_required.
3804 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3805 message was being sent as an SMTP response.
3807 PH/61 Add -l and -k options to exicyclog.
3809 PH/62 When verifying, if an address was redirected to one new address, so that
3810 verification continued, and the new address failed or deferred after
3811 having set something in $address_data, the value of $address_data was not
3812 passed back to the ACL. This was different to the case when no
3813 redirection occurred. The value is now passed back in both cases.
3815 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3816 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3817 use login_cap.h, so on its own it isn't the distinguishing feature. The
3818 new name refers directly to the setclassresources() function.
3820 PH/65 Added configuration files for NetBSD3.
3822 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3824 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3825 is preferred over IPv4.
3827 PH/68 The bounce_return_message and bounce_return_body options were not being
3828 honoured for bounces generated during the reception of non-SMTP messages.
3829 In particular, this applied to messages rejected by the ACL. This bug has
3830 been fixed. However, if bounce_return_message is true and bounce_return_
3831 body is false, the headers that are returned for a non-SMTP message
3832 include only those that have been read before the error was detected.
3833 (In the case of an ACL rejection, they have all been read.)
3835 PH/69 The HTML version of the specification is now built in a directory called
3836 spec_html instead of spec.html, because the latter looks like a path with
3837 a MIME-type, and this confuses some software.
3839 PH/70 Catch two compiler warnings in sieve.c.
3841 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3842 function verify_get_ident() calls ip_connect() to connect a socket, but
3843 if the "connect()" function timed out, ip_connect() used to close the
3844 socket. However, verify_get_ident() also closes the socket later, and in
3845 between Exim writes to the log, which may get opened at this point. When
3846 the socket was closed in ip_connect(), the log could get the same file
3847 descriptor number as the socket. This naturally causes chaos. The fix is
3848 not to close the socket in ip_connect(); the socket should be closed by
3849 the function that creates it. There was only one place in the code where
3850 this was missing, in the iplookup router, which I don't think anybody now
3851 uses, but I've fixed it anyway.
3853 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3854 well as to direct DNS lookups. Otherwise the handling of names in host
3855 lists is inconsistent and therefore confusing.
3861 PH/01 Two changes to the default runtime configuration:
3863 (1) Move the checks for relay_from_hosts and authenticated clients from
3864 after to before the (commented out) DNS black list checks.
3866 (2) Add control=submission to the relay_from_hosts and authenticated
3867 clients checks, on the grounds that messages accepted by these
3868 statements are most likely to be submissions.
3870 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3872 (1) Generate an error if the third argument for the ${prvs expansion is
3875 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3878 (3) Reset the variables that are obtained from the first argument of
3879 ${prvscheck and used in the second argument before leaving the code,
3880 because their memory is reclaimed, so using them afterwards may do
3883 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3884 one (it's much easier than Tom thought :-).
3886 (5) Because of (4), we can now allow for the use of $prvscheck_result
3887 inside the third argument.
3889 PH/03 For some reason, the default setting of PATH when running a command from
3890 a pipe transport was just "/usr/bin". I have changed it to
3893 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3894 anything to be listed in the output from -bV.
3896 PH/05 When a filter generated an autoreply, the entire To: header line was
3897 quoted in the delivery log line, like this:
3899 => >A.N.Other <ano@some.domain> <original@ddress> ...
3901 This has been changed so that it extracts the operative address. There
3902 may be more than one such address. If so, they are comma-separated, like
3905 => >ano@some.domain,ona@other.domain <original@ddress> ...
3907 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3908 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3909 not being looked up in the rDNS to get a host name, Exim was showing the
3910 IP address twice in Received: lines, even though the IP addresses were
3911 identical. For example:
3913 Received: from [1.2.3.4] (helo=[1.2.3.4])
3915 However, if the real host name was known, it was omitting the HELO data
3916 if it matched the actual IP address. This has been tidied up so that it
3917 doesn't show the same IP address twice.
3919 PH/07 When both +timestamp and +memory debugging was on, the value given by
3920 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3921 called by the debug printing, thereby overwriting the timestamp buffer.
3922 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3924 PH/08 When the original message was included in an autoreply transport, it
3925 always said "this is a copy of the message, including all the headers",
3926 even if body_only or headers_only was set. It now gives an appropriate
3929 PH/09 Applied a patch from the Sieve maintainer which:
3931 o fixes some comments
3932 o adds the (disabled) notify extension core
3933 o adds some debug output for the result of if/elsif tests
3934 o points to the current vacation draft in the documentation
3935 and documents the missing references header update
3939 o fixes a bug in processing the envelope test (when testing
3940 multiple envelope elements, the last element determined the
3943 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3944 Electronic Mail") by including:
3946 Auto-submitted: auto-generated
3948 in the messages that it generates (bounce messages and others, such as
3949 warnings). In the case of bounce messages for non-SMTP messages, there was
3950 also a typo: it was using "Auto_submitted" (underscore instead of
3951 hyphen). Since every message generated by Exim is necessarily in response
3952 to another message, thes have all been changed to:
3954 Auto-Submitted: auto-replied
3956 in accordance with these statements in the RFC:
3958 The auto-replied keyword:
3960 - SHOULD be used on messages sent in direct response to another
3961 message by an automatic process,
3963 - MUST NOT be used on manually-generated messages,
3965 - MAY be used on Delivery Status Notifications (DSNs) and Message
3966 Disposition Notifications (MDNs),
3968 - MUST NOT be used on messages generated by automatic or periodic
3969 processes, except for messages which are automatic responses to
3972 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3973 to the default Received: header definition.
3975 PH/12 Added log selector acl_warn_skipped (default on).
3977 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3978 variables because (a) they are in the wrong storage pool and (b) even if
3979 they were copied, it wouldn't work properly because of the caching.
3981 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3982 checking when decoding. Apparently there are clients that generate
3983 overlong encoded strings. Why am I not surprised?
3985 PH/15 If the first argument of "${if match_address" was not empty, but did not
3986 contain an "@" character, Exim crashed. Now it writes a panic log message
3987 and treats the condition as false.
3989 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3991 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3992 extension "envelope-auth". The code is finished and in agreement with
3993 other implementations, but there is no documentation so far and in fact,
3994 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3995 not changing the active code.
3997 Print executed "if" and "elsif" statements when debugging is used. This
3998 helps a great deal to understand what a filter does.
4000 Document more things not specified clearly in RFC3028. I had all this
4001 sorted out, when out of a sudden new issues came to my mind. Oops."
4003 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4006 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4007 canonical form (as documented). However, after a host name lookup from
4008 the IP address, check_host() was doing a simple string comparison with
4009 addresses acquired from the DNS when checking that the found name did
4010 have the original IP as one of its addresses. Since any found IPv6
4011 addresses are likely to be in abbreviated form, the comparison could
4012 fail. Luckily, there already exists a function for doing the comparison
4013 by converting both addresses to binary, so now that is used instead of
4014 the text comparison.
4016 PH/20 There was another similar case to PH/19, when a complete host name was
4017 given in a host list; looking up its IP address could give an abbreviated
4018 form, whereas the current host's name might or might not be abbreviated.
4019 The same fix has been applied.
4025 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4026 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4029 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4030 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4032 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4034 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4035 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4036 Also, for "Auto-Submitted", treat anything other than "no" as
4037 non-personal, in accordance with RFC 3834. (Previously it treated
4038 anything starting "auto-" as non-personal.)
4040 TF/01 The control=submission/name=... option had a problem with syntax
4041 errors if the name included a slash character. The /name= option
4042 now slurps the rest of the string, so it can include any characters
4043 but it must come last in the list of options (after /sender_retain
4046 PH/05 Some modifications to the interface to the fake nameserver for the new
4054 TK/01 Added the "success_on_redirect" address verification option. See
4055 NewStuff for rationale and an example.
4057 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4059 PH/02 Patch to exigrep to allow it to work on syslog lines.
4061 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4062 fread() to skip over the body file's header line, because in Cygwin the
4063 header line is locked and is inaccessible.
4065 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4066 co-exist for some time) to make it clear that it is the Exim ID that is
4067 referenced, not the Message-ID: header line.
4069 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4070 string_format() function, because snprintf() does not exist on all
4073 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4074 ${stat: expansion item.
4076 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4077 protocol synchronization error", to keep the pedants happy.
4079 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4080 well as for IRIX systems, when gcc is being used. See the host.c source
4083 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4085 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4088 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4089 transport and to the smtp transport in LMTP mode.
4091 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4093 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4094 run a filter in a subprocess. This could lead to confusion in subsequent
4095 lookups in the parent process. There should also be a search_tidyup() at
4096 the end of the subprocess.
4098 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4099 only if the host matched helo_try_verify_hosts, which caused the
4100 verification to occur when the EHLO/HELO command was issued. The ACL just
4101 tested the remembered result. Now, if a previous verification attempt has
4102 not happened, "verify = helo" does it there and then.
4104 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4106 TK/03 Fix log output including CR from clamd.
4108 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4109 exist provoked a memory error which could cause a segfault.
4111 PH/15 Installed PCRE 6.2
4113 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4115 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4116 of the problem. Specifically, suggested +O2 rather than +O1 for the
4119 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4121 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4122 record, the port was not correctly logged when the outgoing_port log
4123 selector was set (it logged the transort's default port).
4125 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4126 fallback_hosts, and "hosts" in the smtp transport.
4128 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4129 host errors such as "Connection refused".
4131 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4132 authentication with radiusclient 0.4.9:
4134 - Error returned from rc_read_config was caught wrongly
4135 - Username/password not passed on to radius server due to wrong length.
4137 The presumption is that some radiusclient API changes for 4.51/PH/17
4138 were not taken care of correctly. The code is still untested by me (my
4139 Linux distribution still has 0.3.2 of radiusclient), but it was
4140 contributed by a Radius user.
4142 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4143 expanding the "port" option of the smtp transport.
4145 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4146 while reading a MIME header. Thanks to Tom Hughes for a patch.
4148 PH/24 Include config.h inside local_scan.h so that configuration settings are
4151 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4152 that in an ACL for RCPT (for example), you can examine exactly what was
4155 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4156 commands, but it was not correctly comparing the address with the actual
4157 client host address. Thus, it would show the EHLO address in Received:
4158 header lines when this was not necessary.
4160 PH/27 Added the % operator to ${eval:}.
4162 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4163 it should be ignoring failures (because with -C, for example, it has lost
4164 privilege). It wasn't ignoring creation failures other than "already
4167 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4170 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4171 because an input line was too long, either on its own, or by virtue of
4172 too many continuations, the temporary file was not being removed, and the
4173 return code was incorrect.
4175 PH/31 Missing "BOOL" in function definition in filtertest.c.
4177 PH/32 Applied Sieve patches from the maintainer.
4179 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4181 PH/33 Added "verify = not_blind".
4183 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4184 Local/Makefile (with some defaults set). These are used in built scripts
4185 such as exicyclog, but they have never been used in the exim_install
4186 script (though there are many overriding facilities there). I have
4187 arranged that the exim_install script now takes note of these two
4190 PH/35 Installed configuration files for Dragonfly.
4192 PH/36 When a locally submitted message by a trusted user did not contain a
4193 From: header, and the sender address was obtained from -f or from an SMTP
4194 MAIL command, and the trusted user did not use -F to supply a sender
4195 name, $originator_name was incorrectly used when constructing a From:
4196 header. Furthermore, $originator_name was used for submission mode
4197 messages from external hosts without From: headers in a similar way,
4198 which is clearly wrong.
4200 PH/37 Added control=suppress_local_fixups.
4202 PH/38 When log_selector = +received_sender was set, and the addition of the
4203 sender made the log line's construction buffer exactly full, or one byte
4204 less than full, an overflow happened when the terminating "\n" was
4207 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4208 when the result of a list match is failure because a DNS lookup failed.
4210 PH/40 RM_COMMAND is now used in the building process.
4212 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4213 the "build-* directories that it finds.
4215 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4216 address in a domain literal was a prefix of an interface address.
4218 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4219 when verifying a sender address, unless rewrite_headers is false.
4221 PH/44 Wrote a long comment about why errors_to addresses are verified as
4222 recipients, not senders.
4224 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4225 the ratelimit ACL was added.
4227 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4229 PH/46 Added extra information about PostgreSQL errors to the error string.
4231 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4232 avoiding the need to install special zones in a real server. This is
4233 backwards compatible; if it can't find the fake resolver, it drops back.
4234 Thus, both old and new test suites can be run.
4236 TF/02 Added util/ratelimit.pl
4238 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4239 clock is set back in time.
4241 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4242 Candler <B.Candler@pobox.com>.
4244 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4245 OK for addresses that are the result of redirections.
4247 PH/48 A number of further additions for the benefit of the new test suite,
4248 including a fake gethostbyname() that interfaces to the fake DNS resolver
4251 TF/06 The fix for widen_domains has also been applied to qualify_single and
4252 search_parents which are the other dnslookup options that can cause
4255 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4258 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4260 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4261 service" if it encountered an MX record with an empty target hostname.
4262 The message is now "an MX or SRV record indicated no SMTP service".
4264 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4265 if the DNS of the sending site is misconfigured. This is quite a
4266 common situation. This change restores the behaviour of treating a
4267 helo verification defer as a failure.
4269 PH/51 If self=fail was set on a router, the bounce message did not include the
4270 actual error message.
4276 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4278 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4279 command itself ended in error, the underlying message about the transport
4280 filter timeout was being overwritten with the pipe command error. Now the
4281 underlying error message should be appended to the second error message.
4283 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4285 PH/02 Reduce the amount of output that "make" produces by default. Full output
4286 can still be requested.
4288 PH/03 The warning log line about a condition test deferring for a "warn" verb
4289 was being output only once per connection, rather than after each
4290 occurrence (because it was using the same function as for successful
4291 "warn" verbs). This seems wrong, so I have changed it.
4293 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4294 it should not have, which might have caused a crash in the right
4295 circumstances, but probably never did.
4297 PH/04 Installed a modified version of Tony Finch's patch to make submission
4298 mode fix the return path as well as the Sender: header line, and to
4299 add a /name= option so that you can make the user's friendly name appear
4302 TF/03 Added the control = fakedefer ACL modifier.
4304 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4305 Mark Lowes for thorough testing.
4307 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4309 TK/03 Merged latest SRS patch from Miles Wilton.
4311 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4312 with the definition in sysexits.h (which is #included earlier).
4313 Fortunately, Exim does not actually use EX_OK. The code used to try to
4314 preserve the sysexits.h value, by assuming that macro definitions were
4315 scanned for macro replacements. I have been disabused of this notion,
4316 so now the code just undefines EX_OK before #including unistd.h.
4318 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4319 in the smtp transport. When a block could not be written in a single
4320 write() function, the timeout was being re-applied to each part-write.
4321 This seems wrong - if the receiver was accepting one byte at a time it
4322 would take for ever. The timeout is now adjusted when this happens. It
4323 doesn't have to be particularly precise.
4325 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4326 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4328 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4329 without a domain if the check to <postmaster@domain> fails.
4331 SC/01 Eximstats: added -xls and the ability to specify output files
4332 (patch written by Frank Heydlauf).
4334 SC/02 Eximstats: use FileHandles for outputting results.
4336 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4338 SC/04 Eximstats: fixed display of large numbers with -nvr option
4340 SC/05 Eximstats: fixed merging of reports with empty tables.
4342 SC/06 Eximstats: added the -include_original_destination flag
4344 SC/07 Eximstats: removed tabs and trailing whitespace.
4346 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4348 TK/06 MBOX spool code: Add real "From " MBOX separator line
4349 so the .eml file is really in mbox format (even though
4350 most programs do not really care). Patch from Alex Miller.
4352 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4353 The latter is generated from $received_to and is only set if the
4354 message has one envelope recipient. SA can use these headers,
4355 obviously out-of-the-box. Patch from Alex Miller.
4357 PH/08 The ${def test on a variable was returning false if the variable's
4358 value was "0", contrary to what the specification has always said!
4359 The result should be true unless the variable is empty.
4361 PH/09 The syntax error of a character other than { following "${if
4362 def:variable_name" (after optional whitespace) was not being diagnosed.
4363 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4364 accidental colon was present, for example, could give incorrect results.
4366 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4367 result is used (not including appendfile, where other changes are about
4370 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4371 This involved changing a lot of size variables from int to off_t. It
4372 should work with maildirs and everything.
4374 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4375 spamd dying while we are connected to it.
4377 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4380 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4381 function for BDB 4.3.
4383 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4385 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4386 the sources for PCRE have been re-arranged and more files are now
4389 PH/15 The code I had for printing potentially long long variables in PH/11
4390 above was not the best (it lost precision). The length of off_t variables
4391 is now inspected at build time, and an appropriate printing format (%ld
4392 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4393 to be "long long int" or "long int". This is needed for the internal
4394 formatting function string_vformat().
4396 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4397 the configuration file to be ":syslog", then the script "guesses" where
4398 the logs files are, rather than using the compiled in default. In our
4399 case the guess is not the same as the compiled default, so the script
4400 suddenly stopped working when I started to use syslog. The patch checks
4401 to see if log_file_path is "". If so, it attempts to read it from exim
4402 with no configuration file to get the compiled in version, before it
4403 falls back to the previous guessing code."
4405 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4406 implementing BATV in an Exim configuration. See NewStuff for the gory
4409 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4410 Makefile that are specific to HP-UX.
4412 PH/18 If the "use_postmaster" option was set for a recipient callout together
4413 with the "random" option, the postmaster address was used as the MAIL
4414 FROM address for the random test, but not for the subsequent recipient
4415 test. It is now used for both.
4417 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4418 patch removes a few documentation additions to RFC 3028, because the
4419 latest draft now contains them. It adds the new en;ascii-case comparator
4420 and a new error check for 8bit text in MIME parts. Comparator and
4421 require names are now matched exactly. I enabled the subaddress
4422 extension, but it is not well tested yet (read: it works for me)."
4424 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4425 rework some of the code of TK/09 above to avoid the hardwired use of
4426 "%lld" and "long long". Replaced the call to snprintf() with a call to
4429 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4430 records point to non-existent hosts", "retry timeout exceeded", and
4431 "retry time not reached for any host after a long failure period".
4433 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4434 experimental DomainKeys support:
4436 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4437 (2) On an error such as an illegally used "control", the wrong name for
4438 the control was given.
4440 These problems did NOT occur unless DomainKeys support was compiled.
4442 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4444 PH/24 Added ${if match_ip condition.
4446 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4447 databases so that it will be absolutely obvious if a crash occurs in the
4448 DB library. This is a regular occurrence (often caused by mis-matched
4451 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4452 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4453 picked up on a user's system that detects such things. There doesn't seem
4454 to be a gcc warning option for this - only an attribute that has to be
4455 put on the function's prototype. It seems that in Fedora Core 4 they have
4456 set this on a number of new functions. No doubt there will be more in due
4459 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4460 specify a transport. However, if an address that was verified by such a
4461 router was the subject of a callout, Exim crashed because it tried to
4462 read the rcpt_include_affixes from the non-existent transport. Now it
4463 just assumes that the setting of that option is false. This bug was
4464 introduced by 4.51/PH/31.
4466 PH/28 Changed -d+all to exclude +memory, because that information is very
4467 rarely of interest, but it makes the output a lot bigger. People tend to
4468 do -d+all out of habit.
4470 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4471 code in os-type was giving problems when libc.so lives in lib64, like on
4474 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4475 aren't the modern standard, and it seems that some systems' include files
4476 don't always have them. Exim was already checking for some of the newer
4477 ones like T_AAAA, and defining it itself. I've added checks for all the
4478 record types that Exim uses.
4480 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4481 not automatically generating a new one, as it is supposed to. This
4482 prevented TLS from working. If the file did exist, but contained invalid
4483 data, a new version was generated, as expected. It was only the case of a
4484 non-existent file that was broken.
4486 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4487 with a change in libdomainkeys > 0.64.
4489 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4490 from DNS. If the selector record carries the flag, it now has
4491 precedence over the domain-wide flag.
4493 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4495 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4496 the use of a port name that isn't defined in /etc/services) occurred, the
4497 message was deferred as in a normal delivery, and thus remained on the
4498 spool, instead of being failed because of the mua_wrapper setting. This
4499 is now fixed, and I tidied up some of the mua_wrapper messages at the
4502 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4503 the messages in a hash of arrays rather than using individual hashes.
4504 This is a bit cleaner and results in dramatic memory savings, albeit
4505 at a slight CPU cost.
4507 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4508 as requested by Marc Sherman.
4510 SC/10 Eximstats: added histograms for user specified patterns as requested
4513 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4515 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4516 fopen() in the content-scanning modules that did not already have it.
4522 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4523 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4525 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4527 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4529 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4530 to test Sieve filters that use "vacation".
4532 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4533 that changes the way the GnuTLS parameters are stored in the cache file.
4534 The new format can be generated externally. For backward compatibility,
4535 if the data in the cache doesn't make sense, Exim assumes it has read an
4536 old-format file, and it generates new data and writes a new file. This
4537 means that you can't go back to an older release without removing the
4540 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4541 work if there are any delivery delays because "one_time" forces the
4542 parent to be marked "delivered", so its unseen clone is never tried
4543 again. For this reason, Exim now forbids the simultaneous setting of
4546 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4547 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4548 entry if you want to know the details. The fix, however, neglected to
4549 consider the case when local delivery batching is involved. The test for
4550 "previously delivered" was not happening when checking to see if an
4551 address could be batched with a previous (undelivered) one; under
4552 certain circumstances this could lead to multiple deliveries to the same
4555 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4556 in its include files, and this causes problems building Exim.
4558 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4559 header_syntax) but Exim was just ignoring anything given after a slash.
4560 In particular, this caused confusion with an attempt to use "verify =
4561 reverse_host_lookup/defer_ok". An error is now given when options are
4562 supplied for verify items that do not have them. (Maybe reverse_host_
4563 lookup should have a defer_ok option, but that's a different point.)
4565 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4566 defined by RFC 821) to 2048, because there were problems with some AUTH
4567 commands, and RFC 1869 says the size should be increased for extended
4568 SMTP commands that take arguments.
4570 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4573 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4574 "unknown" error; now it says that the functionality isn't in the binary.
4576 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4577 an address' error message when a string expansion fails (syntax or
4578 whatever). Otherwise the password may appear in the log. Following change
4579 PH/42 below, there is no longer a chance of it appearing in a bounce
4582 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4584 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4585 of its addresses was ever tried. (Bugzilla bug #2.)
4587 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4588 the result incorrectly in the debug output. (It correctly added a newline
4589 to what was transported.)
4591 TF/01 Added $received_time.
4593 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4594 commented out examples of how to interface to a virus scanner and to
4595 SpamAssassin. Also added commented examples of av_scanner and
4596 spamd_address settings.
4598 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4599 and controls are allowed in which ACLs. There were a couple of minor
4600 errors. Some of the entries in the conditions table (which is a table of
4601 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4602 negation of where the condition IS allowed.
4604 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4606 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4607 header file does not have a version number, so I've had to invent a new
4608 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4609 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4610 radiusclient), but it was contributed by a Radius user.
4612 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4613 files or directories, for OpenSSL.
4615 PH/19 When an Exim process that is running as root has to create an Exim log
4616 file, it does so in a subprocess that runs as exim:exim so as to get the
4617 ownership right at creation (otherwise, other Exim processes might see
4618 the file with the wrong ownership). There was no test for failure of this
4619 fork() call, which would lead to the process getting stuck as it waited
4620 for a non-existent subprocess. Forks do occasionally fail when resources
4621 run out. I reviewed all the other calls to fork(); they all seem to check
4624 PH/20 When checking for unexpected SMTP input at connect time (before writing
4625 the banner), Exim was not dealing correctly with a non-positive return
4626 from the read() function. If the client had disconnected by this time,
4627 the result was a log entry for a synchronization error with an empty
4628 string after "input=" when read() returned zero. If read() returned -1
4629 (an event I could not check), uninitialized data bytes were printed.
4630 There were reports of junk text (parts of files, etc) appearing after
4633 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4635 PH/22 Added support for macro redefinition, and (re)definition in between
4636 driver and ACL definitions.
4638 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4639 forgetting to use the resulting value; it was using the unexpanded value.
4641 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4642 hadn't been configured. The fix is from Juergen Kreileder, who
4643 understands it better than I do:
4645 "Here's what I see happening with three configured cyrus_sasl
4646 authenticators configured (plain, login, cram-md5):
4648 On startup auth_cyrus_sasl_init() gets called for each of these.
4649 This means three calls to sasl_listmech() without a specified mech_list.
4650 => SASL tests which mechs of all available mechs actually work
4651 => three warnings about OTP not working
4652 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4654 With the patch, sasl_listmech() also gets called three times. But now
4655 SASL's mech_list option is set to the server_mech specified in the the
4656 authenticator. Or in other words, the answer from sasl_listmech()
4657 gets limited to just the mech you're testing for (which is different
4659 => the return list contains just 'plain' or 'login', 'cram-md5' or
4660 nothing depending on the value of ob->server_mech.
4662 I've just tested the patch: Authentication still works fine,
4663 unavailable mechs specified in the exim configuration are still
4664 caught, and the auth.log warnings about OTP are gone."
4666 PH/25 When debugging is enabled, the contents of the command line are added
4667 to the debugging output, even when log_selector=+arguments is not
4670 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4671 answer is "GNU", and only if the return is "GNU/something" is the answer
4674 PH/27 $acl_verify_message is now set immediately after the failure of a
4675 verification in an ACL, and so is available in subsequent modifiers. In
4676 particular, the message can be preserved by coding like this:
4678 warn !verify = sender
4679 set acl_m0 = $acl_verify_message
4681 Previously, $acl_verify_message was set only while expanding "message"
4682 and "log_message" when a very denied access.
4684 PH/28 Modified OS/os.c-Linux with
4686 -#ifndef OS_LOAD_AVERAGE
4687 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4689 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4690 nomenclature these days.)
4692 PH/29 Installed patch from the Sieve maintainer that adds the options
4693 sieve_useraddress and sieve_subaddress to the redirect router.
4695 PH/30 In these circumstances:
4696 . Two addresses routed to the same list of hosts;
4697 . First host does not offer TLS;
4698 . First host accepts first address;
4699 . First host gives temporary error to second address;
4700 . Second host offers TLS and a TLS session is established;
4701 . Second host accepts second address.
4702 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4703 and peerdn, if requested) that were in fact used only for the second
4706 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4707 attention to any local part prefix or suffix that was matched by the
4708 router that accepted the address. It now behaves in the same way as it
4709 does for delivery: the affixes are removed from the local part unless
4710 rcpt_include_affixes is set on the transport.
4712 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4713 timeout during the DATA phase of an incoming message.
4715 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4716 applied a patch sanctioned by the Sieve maintainer.
4718 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4719 the uid or gid is negative. A case of a negative gid caused this to be
4720 noticed. The fix allows for either to be negative.
4722 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4723 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4726 PH/36 The change PH/12 above was broken. Fixed it.
4728 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4729 the grounds that routing the same address twice would always produce the
4730 same answer. This might have been true once, but it is certainly no
4731 longer true now. Routing a child address may depend on the previous
4732 routing that produced that child. Some complicated redirection strategies
4733 went wrong when messages had multiple recipients, and made Exim's
4734 behaviour dependent on the order in which the addresses were given.
4736 I have moved the duplicate checking until after the routing is complete.
4737 Exim scans the addresses that are assigned to local and remote
4738 transports, and removes any duplicates. This means that more work will be
4739 done, as duplicates will always all be routed, but duplicates are
4740 presumably rare, so I don't expect this is of any significance.
4742 For deliveries to pipes, files, and autoreplies, the duplicate checking
4743 still happens during the routing process, since they are not going to be
4746 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4747 It corrects a timeout issue with spamd. This is Ian's comment: "The
4748 background is that sometimes spamd either never reads data from a
4749 connection it has accepted, or it never writes response data. The exiscan
4750 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4751 blindly assumes that writes won't block so it may never time out."
4753 PH/39 Allow G after quota size as well as K and M.
4755 PH/40 The value set for $authenticated_id in an authenticator may not contain
4756 binary zeroes or newlines because the value is written to log lines and
4757 to spool files. There was no check on this. Now the value is run through
4758 the string_printing() function so that such characters are converted to
4759 printable escape sequences.
4761 PH/41 $message_linecount is a new variable that contains the total number of
4762 lines in the message. Compare $body_linecount, which is the count for the
4765 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4766 bounce and delay warning messages, except in certain special cases, which
4769 (a) An SMTP error message from a remote host;
4770 (b) A message specified in a :fail: redirection;
4771 (c) A message specified in a "fail" command in a system filter;
4772 (d) A message specified in a FAIL return from the queryprogram router;
4773 (e) A message specified by the cannot_route_message router option.
4775 In these cases only, Exim does include the error details in bounce and
4776 warning messages. There are also a few cases where bland messages such
4777 as "unrouteable address" or "local delivery error" are given.
4779 PH/43 $value is now also set for the "else" part of a ${run expansion.
4781 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4782 being worked on, but at least Exim now implements the latest version to
4785 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4786 process to complete was treated as a delivery failure, a timeout while
4787 writing the message to the pipe was logged, but erroneously treated as a
4788 successful delivery. Such timeouts include transport filter timeouts. For
4789 consistency with the overall process timeout, these timeouts are now
4790 treated as errors, giving rise to delivery failures by default. However,
4791 there is now a new Boolean option for the pipe transport called
4792 timeout_defer, which, if set TRUE, converts the failures into defers for
4793 both kinds of timeout. A transport filter timeout is now identified in
4796 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4797 systems where "make" and "gmake" are different, calling "gmake" at top
4798 level broke things. I've arranged for the value of $(MAKE) to be passed
4799 from the Makefile to this script so that it can call the same version of
4803 A note about Exim versions 4.44 and 4.50
4804 ----------------------------------------
4806 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4807 changes of various kinds. As a consequence, a big documentation update was
4808 needed. This delayed the release for rather longer than seemed good, especially
4809 in the light of a couple of (minor) security issues. Therefore, the changes
4810 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4811 release. So 4.44 and 4.50 are in effect two different branches that both start
4814 I have left the 4.50 change log unchanged; it contains all the changes since
4815 4.43. The change log for 4.44 is below; many of its items are identical to
4816 those for 4.50. This seems to be the most sensible way to preserve the
4817 historical information.
4823 1. Minor wording change to the doc/README.SIEVE file.
4825 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4826 computation of the current number of files was incorrect.
4828 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4829 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4832 4. Give more explanation in the error message when the command for a transport
4833 filter fails to execute.
4835 5. There are several places where Exim runs a non-Exim command in a
4836 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4837 was being done only for the command run by the queryprogram router. It is
4838 now done for all such subprocesses. The other cases are: ${run, transport
4839 filters, and the commands run by the lmtp and pipe transports.
4841 6. Added CONFIGURE_GROUP build-time option.
4843 7. Some older OS have a limit of 256 on the maximum number of file
4844 descriptors. Exim was using setrlimit() to set 1000 as a large value
4845 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4846 systems. I've change it so that if it can't get 1000, it tries for 256.
4848 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4849 was an oversight, and furthermore, ever since the addition of extra
4850 controls (e.g. 4.43/32), the checks on when to allow different forms of
4851 "control" were broken. There should now be diagnostics for all cases when a
4852 control that does not make sense is encountered.
4854 9. Added the /retain_sender option to "control=submission".
4856 10. $recipients is now available in the predata ACL (oversight).
4858 11. Tidy the search cache before the fork to do a delivery from a message
4859 received from the command line. Otherwise the child will trigger a lookup
4860 failure and thereby defer the delivery if it tries to use (for example) a
4861 cached ldap connection that the parent has called unbind on.
4863 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4864 of $address_data from the recipient verification was clobbered by the
4865 sender verification.
4867 13. The value of address_data from a sender verification is now available in
4868 $sender_address_data in subsequent conditions in the ACL statement.
4870 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4872 15. Added a new option "connect=<time>" to callout options, to set a different
4875 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4876 was its contents. (It was OK if the option was not defined at all.)
4878 17. A "Completed" log line is now written for messages that are removed from
4879 the spool by the -Mrm option.
4881 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4882 information about exactly what failed.
4884 19. Added -dd to debug only the daemon process.
4886 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4887 handles timeouts, both on the server side and network timeouts. Renamed the
4888 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4890 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4891 It is now set to "smtps".
4893 22. $host_address is now set to the target address during the checking of
4894 ignore_target_hosts.
4896 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4897 being passed; this would have caused $sender_host_name to have been used if
4898 matching the list had actually called for a host name (not very likely,
4899 since this list is usually IP addresses). A host name is now passed as
4902 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4903 code that specifies a non-restarting handler (typically sigaction() in
4904 modern systems) in an attempt to fix a rare and obscure crash bug.
4906 25. Narrowed the window for a race in the daemon that could cause it to ignore
4907 SIGCHLD signals. This is not a major problem, because they are used only to
4908 wake it up if nothing else does.
4910 26. A malformed maildirsize file could cause Exim to calculate negative values
4911 for the mailbox size or file count. Odd effects could occur as a result.
4912 The maildirsize information is now recalculated if the size or filecount
4915 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4916 support for a long time. Removed HAVE_SYS_VFS_H.
4918 28. Installed the latest version of exipick from John Jetmore.
4920 29. In an address list, if the pattern was not a regular expression, an empty
4921 subject address (from a bounce message) matched only if the pattern was an
4922 empty string. Non-empty patterns were not even tested. This was the wrong
4923 because it is perfectly reasonable to use an empty address as part of a
4924 database query. An empty address is now tested by patterns that are
4925 lookups. However, all the other forms of pattern expect the subject to
4926 contain a local part and a domain, and therefore, for them, an empty
4927 address still always fails if the pattern is not itself empty.
4929 30. Exim went into a mad DNS loop when attempting to do a callout where the
4930 host was specified on an smtp transport, and looking it up yielded more
4931 than one IP address.
4933 31. Re-factored the code for checking spool and log partition space into a
4934 function that finds that data and another that does the check. The former
4935 is then used to implement four new variables: $spool_space, $log_space,
4936 $spool_inodes, and $log_inodes.
4938 32. The RFC2047 encoding function was originally intended for short strings
4939 such as real names; it was not keeping to the 75-character limit for
4940 encoded words that the RFC imposes. It now respects the limit, and
4941 generates multiple encoded words if necessary. To be on the safe side, I
4942 have increased the buffer size for the ${rfc2047: expansion operator from
4945 33. It is now permitted to omit both strings after an "if" condition; if the
4946 condition is true, the result is "true". As before, when the second string
4947 is omitted, a false condition yields an empty string. This makes it less
4948 cumbersome to write custom ACL and router conditions.
4950 34. Failure to deliver a bounce message always caused it to be frozen, even if
4951 there was an errors_to setting on the router. The errors_to setting is now
4954 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4955 canonical form (fully expanded) before being placed in
4956 $sender_host_address.
4958 36. The table in the code that translates DNS record types into text (T_A to
4959 "A" for instance) was missing entries for NS and CNAME. It is just possible
4960 that this could have caused confusion if both these types were looked up
4961 for the same domain, because the text type is used as part of Exim's
4962 per-process caching. But the chance of anyone hitting this buglet seems
4965 37. The dnsdb lookup has been extended in a number of ways.
4967 (1) There is a new type, "zns", which walks up the domain tree until it
4968 finds some nameserver records. It should be used with care.
4970 (2) There is a new type, "mxh", which is like "mx" except that it returns
4971 just the host names, not the priorities.
4973 (3) It is now possible to give a list of domains (or IP addresses) to be
4974 looked up. The behaviour when one of the lookups defers can be
4975 controlled by a keyword.
4977 (4) It is now possible to specify the separator character for use when
4978 multiple records are returned.
4980 38. The dnslists ACL condition has been extended: it is now possible to supply
4981 a list of IP addresses and/or domains to be looked up in a particular DNS
4984 39. Added log_selector=+queue_time_overall.
4986 40. When running the queue in the test harness, wait just a tad after forking a
4987 delivery process, to get repeatability of debugging output.
4989 41. Include certificate and key file names in error message when GnuTLS fails
4990 to set them up, because the GnuTLS error message doesn't include the name
4991 of the failing file when there is a problem reading it.
4993 42. Allow both -bf and -bF in the same test run.
4995 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4997 44. The "Exiscan patch" is now merged into the mainline Exim source.
4999 45. Sometimes the final signoff response after QUIT could fail to get
5000 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5001 < 0 before doing a fflush(). This bug looks as though it goes back to the
5002 introduction of TLS in release 3.20, but "sometimes" must have been rare
5003 because the tests only now provoked it.
5005 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5006 (this can affect the format of dates).
5008 47. exim_tidydb, when checking for the continued existence of a message for
5009 which it has found a message-specific retry record, was not finding
5010 messages that were in split spool directories. Consequently, it was
5011 deleting retry records that should have stayed in existence.
5013 48. Steve fixed some bugs in eximstats.
5015 49. The SPA authentication driver was not abandoning authentication and moving
5016 on to the next authenticator when an expansion was forced to fail,
5017 contradicting the general specification for all authenticators. Instead it
5018 was generating a temporary error. It now behaves as specified.
5020 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5021 (the order specifies the preference for clients). The order is now AES256,
5022 AES128, 3DES, ARCFOUR128.
5024 51. Small patch to Sieve code - explicitly set From: when generating an
5027 52. Exim crashed if a remote delivery caused a very long error message to be
5028 recorded - for instance if somebody sent an entire SpamAssassin report back
5029 as a large number of 550 error lines. This bug was coincidentally fixed by
5030 increasing the size of one of Exim's internal buffers (big_buffer) that
5031 happened as part of the Exiscan merge. However, to be on the safe side, I
5032 have made the code more robust (and fixed the comments that describe what
5035 53. Now that there can be additional text after "Completed" in log lines (if
5036 the queue_time_overall log selector is set), a one-byte patch to exigrep
5037 was needed to allow it to recognize "Completed" as not the last thing in
5040 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5041 patch that reportedly fixes this has been added. I am not expert enough to
5042 create a test for it. This is what the patch creator wrote:
5044 "I found a little strange behaviour of ldap code when working with
5045 Windows 2003 AD Domain, where users was placed in more than one
5046 Organization Units. When I tried to give exim partial DN, the exit code
5047 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5048 But simultaneously result of request was absolutely normal ldap result,
5049 so I produce this patch..."
5051 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5052 REFERENCE, so I have modified the code to exclude the patch when that macro
5055 55. Some experimental protocols are using DNS PTR records for new purposes. The
5056 keys for these records are domain names, not reversed IP addresses. The
5057 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5058 leaves it alone. Component reversal etc. now happens only for IP addresses.
5061 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5063 57. Double the size of the debug message buffer (to 2048) so that more of very
5064 long debug lines gets shown.
5066 58. The exicyclog utility now does better if the number of log files to keep
5067 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5069 59. Two changes related to the smtp_active_hostname option:
5071 (1) $smtp_active_hostname is now available as a variable.
5072 (2) The default for smtp_banner uses $smtp_active_hostname instead
5073 of $primary_hostname.
5075 60. The host_aton() function is supposed to be passed a string that is known
5076 to be a valid IP address. However, in the case of IPv6 addresses, it was
5077 not checking this. This is a hostage to fortune. Exim now panics and dies
5078 if the condition is not met. A case was found where this could be provoked
5079 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5080 components; fortuitously, this particular loophole had already been fixed
5081 by change 4.50/55 above.
5083 If there are any other similar loopholes, the new check in host_aton()
5084 itself should stop them being exploited. The report I received stated that
5085 data on the command line could provoke the exploit when Exim was running as
5086 exim, but did not say which command line option was involved. All I could
5087 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5088 running as the user.
5091 61. There was a buffer overflow vulnerability in the SPA authentication code
5092 (which came originally from the Samba project). I have added a test to the
5093 spa_base64_to_bits() function which I hope fixes it.
5096 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5097 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5099 63. The daemon start-up calls getloadavg() while still root for those OS that
5100 need the first call to be done as root, but it missed one case: when
5101 deliver_queue_load_max is set with deliver_drop_privilege. This is
5102 necessary for the benefit of the queue runner, because there is no re-exec
5103 when deliver_drop_privilege is set.
5105 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5106 This has been fixed.
5108 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5109 was in use, was not putting the data itself into the right store pool;
5110 consequently, it could be overwritten for a subsequent message in the same
5111 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5114 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5116 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5117 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5118 the function were treating the return as a boolean value, which happened to
5119 work because 0=false and not-0=true, but is not correct code.
5121 68. The host_aton() function was not handling scoped IPv6 addresses (those
5122 with, for example, "%eth0" on the end) correctly.
5124 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5125 negated items (that is, ~something) in unsigned ints. Some compilers
5126 apparently mutter when there is no cast.
5128 70. If an address verification called from an ACL failed, and did not produce a
5129 user-specific message (i.e. there was only a "system" message), nothing was
5130 put in $acl_verify_message. In this situation, it now puts the system
5133 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5134 session; change 4.31/43 added the unwanted input to the log line - except
5135 that it did not do this in the start of session case. It now does.
5137 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5138 This is wrong and can cause the other end to generate a synchronization
5139 error if it is another Exim or anything else that does the synchronization
5140 check. A QUIT command is no longer sent after a timeout.
5142 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5143 during host lookups.
5145 74. The defer_ok option of callout verification was not working if it was used
5146 when verifying addresses in header lines, that is, for this case:
5148 verify = header_sender/callout=defer_ok
5150 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5151 those file descriptors could be used for SMTP connections. If anything
5152 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5153 could be sent to the SMTP client, causing chaos. The daemon now opens
5154 stdin, stdout, and stderr to /dev/null when it puts itself into the
5157 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5158 log by default. The user can override this with suitable Perl magic.
5160 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5161 the log message when discard triggers, was not working for the DATA ACL or
5162 for the non-SMTP ACL.
5164 78. Error message wording change in sieve.c.
5166 79. If smtp_accept_max_per_host was set, the number of connections could be
5167 restricted to fewer than expected, because the daemon was trying to set up
5168 a new connection before checking whether the processes handling previous
5169 connections had finished. The check for completed processes is now done
5170 earlier. On busy systems, this bug wouldn't be noticed because something
5171 else would have woken the daemon, and it would have reaped the completed
5174 80. If a message was submitted locally by a user whose login name contained one
5175 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5176 It caused a spool format error. I have fixed the spool reading code. A
5177 related problem was that the "from" clause in the Received: line became
5178 illegal because of the space(s). It is now covered by ${quote_local_part.
5180 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5183 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5185 83. Patch from Sieve maintainer to fix unterminated string problem in
5186 "vacation" handling.
5188 84. Some minor changes to the Linux configuration files to help with other
5189 OS variants using glibc.
5191 85. One more patch for Sieve to update vacation handling to latest spec.
5194 ----------------------------------------------------
5195 See the note above about the 4.44 and 4.50 releases.
5196 ----------------------------------------------------
5202 1. Change 4.43/35 introduced a bug that caused file counts to be
5203 incorrectly computed when quota_filecount was set in an appendfile
5206 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5207 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5210 3. Give more explanation in the error message when the command for a transport
5211 filter fails to execute.
5213 4. There are several places where Exim runs a non-Exim command in a
5214 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5215 was being done only for the command run by the queryprogram router. It is
5216 now done for all such subprocesses. The other cases are: ${run, transport
5217 filters, and the commands run by the lmtp and pipe transports.
5219 5. Some older OS have a limit of 256 on the maximum number of file
5220 descriptors. Exim was using setrlimit() to set 1000 as a large value
5221 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5222 systems. I've change it so that if it can't get 1000, it tries for 256.
5224 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5225 was an oversight, and furthermore, ever since the addition of extra
5226 controls (e.g. 4.43/32), the checks on when to allow different forms of
5227 "control" were broken. There should now be diagnostics for all cases when a
5228 control that does not make sense is encountered.
5230 7. $recipients is now available in the predata ACL (oversight).
5232 8. Tidy the search cache before the fork to do a delivery from a message
5233 received from the command line. Otherwise the child will trigger a lookup
5234 failure and thereby defer the delivery if it tries to use (for example) a
5235 cached ldap connection that the parent has called unbind on.
5237 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5238 of $address_data from the recipient verification was clobbered by the
5239 sender verification.
5241 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5242 was its contents. (It was OK if the option was not defined at all.)
5244 11. A "Completed" log line is now written for messages that are removed from
5245 the spool by the -Mrm option.
5247 12. $host_address is now set to the target address during the checking of
5248 ignore_target_hosts.
5250 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5251 being passed; this would have caused $sender_host_name to have been used if
5252 matching the list had actually called for a host name (not very likely,
5253 since this list is usually IP addresses). A host name is now passed as
5256 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5257 code that specifies a non-restarting handler (typically sigaction() in
5258 modern systems) in an attempt to fix a rare and obscure crash bug.
5260 15. Narrowed the window for a race in the daemon that could cause it to ignore
5261 SIGCHLD signals. This is not a major problem, because they are used only to
5262 wake it up if nothing else does.
5264 16. A malformed maildirsize file could cause Exim to calculate negative values
5265 for the mailbox size or file count. Odd effects could occur as a result.
5266 The maildirsize information is now recalculated if the size or filecount
5269 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5270 support for a long time. Removed HAVE_SYS_VFS_H.
5272 18. Updated exipick to current release from John Jetmore.
5274 19. Allow an empty sender to be matched against a lookup in an address list.
5275 Previously the only cases considered were a regular expression, or an
5278 20. Exim went into a mad DNS lookup loop when doing a callout where the
5279 host was specified on the transport, if the DNS lookup yielded more than
5282 21. The RFC2047 encoding function was originally intended for short strings
5283 such as real names; it was not keeping to the 75-character limit for
5284 encoded words that the RFC imposes. It now respects the limit, and
5285 generates multiple encoded words if necessary. To be on the safe side, I
5286 have increased the buffer size for the ${rfc2047: expansion operator from
5289 22. Failure to deliver a bounce message always caused it to be frozen, even if
5290 there was an errors_to setting on the router. The errors_to setting is now
5293 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5294 canonical form (fully expanded) before being placed in
5295 $sender_host_address.
5297 24. Updated eximstats to version 1.33
5299 25. Include certificate and key file names in error message when GnuTLS fails
5300 to set them up, because the GnuTLS error message doesn't include the name
5301 of the failing file when there is a problem reading it.
5303 26. Expand error message when OpenSSL has problems setting up cert/key files.
5306 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5307 (this can affect the format of dates).
5309 28. exim_tidydb, when checking for the continued existence of a message for
5310 which it has found a message-specific retry record, was not finding
5311 messages that were in split spool directories. Consequently, it was
5312 deleting retry records that should have stayed in existence.
5314 29. eximstats updated to version 1.35
5315 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5316 1.35 - bugfix such that pie charts by volume are generated correctly
5318 30. The SPA authentication driver was not abandoning authentication and moving
5319 on to the next authenticator when an expansion was forced to fail,
5320 contradicting the general specification for all authenticators. Instead it
5321 was generating a temporary error. It now behaves as specified.
5323 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5324 (the order specifies the preference for clients). The order is now AES256,
5325 AES128, 3DES, ARCFOUR128.
5327 31. Small patch to Sieve code - explicitly set From: when generating an
5330 32. Exim crashed if a remote delivery caused a very long error message to be
5331 recorded - for instance if somebody sent an entire SpamAssassin report back
5332 as a large number of 550 error lines. This bug was coincidentally fixed by
5333 increasing the size of one of Exim's internal buffers (big_buffer) that
5334 happened as part of the Exiscan merge. However, to be on the safe side, I
5335 have made the code more robust (and fixed the comments that describe what
5338 33. Some experimental protocols are using DNS PTR records for new purposes. The
5339 keys for these records are domain names, not reversed IP addresses. The
5340 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5341 leaves it alone. Component reversal etc. now happens only for IP addresses.
5344 34. The host_aton() function is supposed to be passed a string that is known
5345 to be a valid IP address. However, in the case of IPv6 addresses, it was
5346 not checking this. This is a hostage to fortune. Exim now panics and dies
5347 if the condition is not met. A case was found where this could be provoked
5348 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5349 components; fortuitously, this particular loophole had already been fixed
5350 by change 4.50/55 or 4.44/33 above.
5352 If there are any other similar loopholes, the new check in host_aton()
5353 itself should stop them being exploited. The report I received stated that
5354 data on the command line could provoke the exploit when Exim was running as
5355 exim, but did not say which command line option was involved. All I could
5356 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5357 running as the user.
5360 35. There was a buffer overflow vulnerability in the SPA authentication code
5361 (which came originally from the Samba project). I have added a test to the
5362 spa_base64_to_bits() function which I hope fixes it.
5365 36. The daemon start-up calls getloadavg() while still root for those OS that
5366 need the first call to be done as root, but it missed one case: when
5367 deliver_queue_load_max is set with deliver_drop_privilege. This is
5368 necessary for the benefit of the queue runner, because there is no re-exec
5369 when deliver_drop_privilege is set.
5371 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5372 was in use, was not putting the data itself into the right store pool;
5373 consequently, it could be overwritten for a subsequent message in the same
5374 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5377 38. Sometimes the final signoff response after QUIT could fail to get
5378 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5379 < 0 before doing a fflush(). This bug looks as though it goes back to the
5380 introduction of TLS in release 3.20, but "sometimes" must have been rare
5381 because the tests only now provoked it.
5387 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5388 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5389 Now it returns an integer. A number of calls were still expecting a T/F
5390 return. Fortuitously, in all cases, the tests worked in OK situations,
5391 which is the norm. However, things would have gone wrong on any write
5392 failures on the smtp file descriptor. This function is used when sending
5393 messages over SMTP and also when doing verify callouts.
5395 2. When Exim is called to do synchronous delivery of a locally submitted
5396 message (the -odf or -odi options), it no longer closes stderr before doing
5399 3. Implemented the mua_wrapper option.
5401 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5403 5. Implemented the functions header_remove(), header_testname(),
5404 header_add_at_position(), and receive_remove_recipient(), and exported them
5407 6. If an ACL "warn" statement specified the addition of headers, Exim already
5408 inserted X-ACL-Warn: at the start if there was no header name. However, it
5409 was not making this test for the second and subsequent header lines if
5410 there were newlines in the string. This meant that an invalid header could
5411 be inserted if Exim was badly configured.
5413 7. Allow an ACL "warn" statement to add header lines at the start or after all
5414 the Received: headers, as well as at the end.
5416 8. Added the rcpt_4xx retry error code.
5418 9. Added postmaster_mailfrom=xxx to callout verification option.
5420 10. Added mailfrom=xxxx to the callout verification option, for verify=
5423 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5424 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5426 12. Inserted some casts to stop certain compilers warning when using pointer
5427 differences as field lengths or precisions in printf-type calls (mostly
5428 affecting debugging statements).
5430 13. Added optional readline() support for -be (dynamically loaded).
5432 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5433 same clock tick as a message's arrival, so that its received time was the
5434 same as the "first fail" time on the retry record, and that message
5435 remained on the queue past the ultimate address timeout, every queue runner
5436 would try a delivery (because it was past the ultimate address timeout) but
5437 after another failure, the ultimate address timeout, which should have then
5438 bounced the address, did not kick in. This was a "< instead of <=" error;
5439 in most cases the first failure would have been in the next clock tick
5440 after the received time, and all would be well.
5442 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5443 being recognized when the domain list was tested by the match_domain
5444 condition in an expansion string.
5446 16. Added the ${str2b64: operator.
5448 17. Exim was always calling setrlimit() to set a large limit for the number of
5449 processes, without checking whether the existing limit was already
5450 adequate. (It did check for the limit on file descriptors.) Furthermore,
5451 errors from getrlimit() and setrlimit() were being ignored. Now they are
5452 logged to the main and panic logs, but Exim does carry on, to try to do its
5453 job under whatever limits there are.
5455 18. Imported PCRE 5.0.
5457 19. Trivial typo in log message " temporarily refused connection" (the leading
5460 20. If the log selector return_path_on_delivery was set and an address was
5461 redirected to /dev/null, the delivery process crashed because it assumed
5462 that a return path would always be set for a "successful" delivery. In this
5463 case, the whole delivery is bypassed as an optimization, and therefore no
5466 21. Internal re-arrangement: the function for sending a challenge and reading
5467 a response while authentication was assuming a zero-terminated challenge
5468 string. It's now changed to take a pointer and a length, to allow for
5469 binary data in such strings.
5471 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5473 23. Exim was not respecting finduser_retries when seeking the login of the
5474 uid under which it was called; it was always trying 10 times. (The default
5475 setting of finduser_retries is zero.) Also, it was sleeping after the final
5476 failure, which is pointless.
5478 24. Implemented tls_on_connect_ports.
5480 25. Implemented acl_smtp_predata.
5482 26. If the domain in control=submission is set empty, Exim assumes that the
5483 authenticated id is a complete email address when it generates From: or
5484 Sender: header lines.
5486 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5487 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5488 chown and chgrp in /bin and hostname in /usr/bin.
5490 28. Exim was keeping the "process log" file open after each use, just as it
5491 does for the main log. This opens the possibility of it remaining open for
5492 long periods when the USR1 signal hits a daemon. Occasional processlog
5493 errors were reported, that could have been caused by this. Anyway, it seems
5494 much more sensible not to leave this file open at all, so that is what now
5497 29. The long-running daemon process does not normally write to the log once it
5498 has entered its main loop, and it closes the log before doing so. This is
5499 so that log files can straightforwardly be renamed and moved. However,
5500 there are a couple of unusual error situations where the daemon does write
5501 log entries, and I had neglected to close the log afterwards.
5503 30. The text of an SMTP error response that was received during a remote
5504 delivery was being truncated at 512 bytes. This is too short for some of
5505 the long messages that one sometimes sees. I've increased the limit to
5508 31. It is now possible to make retry rules that apply only when a message has a
5509 specific sender, in particular, an empty sender.
5511 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5512 it possible to be selective about when SMTP synchronization is enforced.
5514 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5516 32. Implemented hosts_connection_nolog.
5518 33. Added an ACL for QUIT.
5520 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5523 35. Added mailbox_size and mailbox_filecount to appendfile.
5525 36. Added control = no_multiline_responses to ACLs.
5527 37. There was a bug in the logic of the code that waits for the clock to tick
5528 in the case where the clock went backwards by a substantial amount such
5529 that the microsecond fraction of "now" was more than the microsecond
5530 fraction of "then" (but the whole seconds number was less).
5532 38. Added support for the libradius Radius client library this is found on
5533 FreeBSD (previously only the radiusclient library was supported).
5539 1. When certain lookups returned multiple values in the form name=value, the
5540 quoting of the values was not always being done properly. Specifically:
5541 (a) If the value started with a double quote, but contained no whitespace,
5543 (b) If the value contained whitespace other than a space character (i.e.
5544 tabs or newlines or carriage returns) it was not quoted.
5545 This fix has been applied to the mysql and pgsql lookups by writing a
5546 separate quoting function and calling it from the lookup code. The fix
5547 should probably also be applied to nisplus, ibase and oracle lookups, but
5548 since I cannot test any of those, I have not disturbed their existing code.
5550 2. A hit in the callout cache for a specific address caused a log line with no
5551 reason for rejecting RCPT. Now it says "Previous (cached) callout
5552 verification failure".
5554 3. There was an off-by-one bug in the queryprogram router. An over-long
5555 return line was truncated at 256 instead of 255 characters, thereby
5556 overflowing its buffer with the terminating zero. As well as fixing this, I
5557 have increased the buffer size to 1024 (and made a note to document this).
5559 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5560 when Exim is waiting for an SMTP response from a remote server, Exim
5561 restarts its select() call on the socket, thereby resetting its timeout.
5562 This is not a problem when such interrupts are rare. Somebody set up a cron
5563 job to run exiwhat every 2 minutes, which is less than the normal select()
5564 timeout (5 or 10 minutes). This meant that the select() timeout never
5565 kicked in because it was always reset. I have fixed this by comparing the
5566 time when an interrupt arrives with the time at the start of the first call
5567 to select(). If more time than the timeout has elapsed, the interrupt is
5568 treated as a timeout.
5570 5. Some internal re-factoring in preparation for the addition of Sieve
5571 extensions (by MH). In particular, the "personal" test is moved to a
5572 separate function, and given an option for scanning Cc: and Bcc: (which is
5573 not set for Exim filters).
5575 6. When Exim created an email address using the login of the caller as the
5576 local part (e.g. when creating a From: or Sender: header line), it was not
5577 quoting the local part when it contained special characters such as @.
5579 7. Installed new OpenBSD configuration files.
5581 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5582 try to make them clearer.
5584 9. Callout options, other than the timeout value, were being ignored when
5585 verifying sender addresses in header lines. For example, when using
5587 verify = header_sender/callout=no_cache
5589 the cache was (incorrectly) being used.
5591 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5592 only the Cygwin environment.
5594 11. When return_path_on_delivery was set as a log selector, if different remote
5595 addresses in the same message used different return paths and parallel
5596 remote delivery occurred, the wrong values would sometimes be logged.
5597 (Whenever a remote delivery process finished, the return path value from
5598 the most recently started remote delivery process was logged.)
5600 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5601 lines when AUTH and/or TLS are in use. This is the "received protocol"
5602 field. Exim used to use "asmtp" for authenticated SMTP, without any
5603 indication (in the protocol name) for TLS use. Now it follows the RFC and
5604 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5605 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5606 names appear in log lines as well as in Received: header lines.
5608 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5609 extensions, and comparison tests, and to fix some bugs.
5611 14. Changes to the "personal" filter test:
5613 (1) The test was buggy in that it was just doing the equivalent of
5614 "contains" tests on header lines. For example, if a user's address was
5615 anne@some.where, the "personal" test would incorrectly be true for
5617 To: susanne@some.where
5619 This test is now done by extracting each address from the header in turn,
5620 and checking the entire address. Other tests that are part of "personal"
5621 are now done using regular expressions (for example, to check local parts
5622 of addresses in From: header lines).
5624 (2) The list of non-personal local parts in From: addresses has been
5625 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5626 taken from the Sieve specification recommendations.
5628 (3) If the message contains any header line starting with "List-" it is
5629 treated as non-personal.
5631 (4) The test for "circular" in the Subject: header line has been removed
5632 because it now seems ill-conceived.
5634 15. Minor typos in src/EDITME comments corrected.
5636 16. Installed latest exipick from John Jetmore.
5638 17. If headers_add on a router specified a text string that was too long for
5639 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5640 of string_sprintf() is now avoided.
5642 18. $message_body_size was not set (it was always zero) when running the DATA
5643 ACL and the local_scan() function.
5645 19. For the "mail" command in an Exim filter, no default was being set for
5646 the once_repeat time, causing a random time value to be used if "once" was
5647 specified. (If the value happened to be <= 0, no repeat happened.) The
5648 default is now 0s, meaning "never repeat". The "vacation" command was OK
5649 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5650 (I found it when inspecting the code).
5652 20. There is now an overall timeout for performing a callout verification. It
5653 defaults to 4 times the callout timeout, which applies to individual SMTP
5654 commands during the callout. The overall timeout applies when there is more
5655 than one host that can be tried. The timeout is checked before trying the
5656 next host. This prevents very long delays if there are a large number of
5657 hosts and all are timing out (e.g. when the network connections are timing
5658 out). The value of the overall timeout can be changed by specifying an
5659 additional sub-option for "callout", called "maxwait". For example:
5661 verify = sender/callout=5s,maxwait=20s
5663 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5664 to the end before writing, but this should make it even safer).
5666 22. Exim was forgetting that it had advertised PIPELINING for the second and
5667 subsequent messages on an SMTP connection. It was also not resetting its
5668 memory on STARTTLS and an internal HELO.
5670 23. When Exim logs an SMTP synchronization error within a session, it now
5671 records whether PIPELINING has been advertised or not.
5673 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5674 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5675 rather than long int.
5677 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5679 26. Added the never_mail option to autoreply.
5685 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5686 crash if the getsockname() call failed; this can happen if a connection is
5687 closed very soon after it is established. The problem was simply in the
5688 order in which certain operations were done, causing Exim to try to write
5689 to the SMTP stream before it had set up the file descriptor. The bug has
5690 been fixed by making things happen in the correct order.
5696 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5697 before the connection was closed, thus losing the rejection response.
5699 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5700 some early Solaris releases, but causes trouble in current releases where
5701 socklen_t is defined.
5703 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5706 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5709 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5710 that releasing the top of it at the end releases what was used for sub-
5711 expansions (unless the block got too big). However, discard this block if
5712 the first thing is a variable or header, so that we can use its block when
5713 it is dynamic (useful for very large $message_headers, for example).
5715 6. Lookups now cache *every* query, not just the most recent. A new, separate
5716 store pool is used for this. It can be recovered when all lookup caches are
5717 flushed. Lookups now release memory at the end of their result strings.
5718 This has involved some general refactoring of the lookup sources.
5720 7. Some code has been added to the store_xxx() functions to reduce the amount
5721 of flapping under certain conditions.
5723 8. log_incoming_interface used to affect only the <= reception log lines. Now
5724 it causes the local interface and port to be added to several more SMTP log
5725 lines, for example "SMTP connection from", and rejection lines.
5727 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5729 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5731 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5732 could be overwritten at the end of the current message (or the start of a
5733 new message if it was set in a HELO ACL). The value is now preserved for
5734 the duration of the SMTP connection.
5736 12. If a transport had a headers_rewrite setting, and a matching header line
5737 contained an unqualified address, that address was qualified, even if it
5738 did not match any rewriting rules. The underlying bug was that the values
5739 of the flags that permit the existence of unqualified sender and recipient
5740 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5741 non-local messages, and by -bnq for local messages) were not being
5742 preserved with the message after it was received.
5744 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5745 "next input=" as part of the text comprising the host identity instead of
5746 the correct text. The code was using the same buffer for two different
5747 strings. However, depending on which order the printing function evaluated
5748 its arguments, the bug did not always show up. Under Linux, for example, my
5749 test suite worked just fine.
5751 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5752 doesn't work with some older versions of Perl. It has been changed to "my",
5753 which in any case is probably the better facility to use.
5755 15. A really picky compiler found some instances of statements for creating
5756 error messages that either had too many or two few arguments for the format
5759 16. The size of the buffer for calls to the DNS resolver has been increased
5760 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5761 for addresses that have a lot of PTR records. This alleviates a problem; it
5762 does not fully solve it.
5764 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5765 buffer now truncates the list and logs the incident, which is the same
5766 action as happens when Exim is looking up a host name and its aliases.
5767 Previously in this situation something unpredictable would happen;
5768 sometimes it was "internal error: store_reset failed".
5770 18. If a server dropped the connection unexpectedly when an Exim client was
5771 using GnuTLS and trying to read a response, the client delivery process
5772 crashed while trying to generate an error log message.
5774 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5775 string, for example:
5777 warn message = H1: something\nH2: something
5779 the text was added as a single header line from Exim's point of view
5780 though it ended up OK in the delivered message. However, searching for the
5781 second and subsequent header lines using $h_h2: did not work. This has been
5782 fixed. Similarly, if a system filter added multiple headers in this way,
5783 the routers could not see them.
5785 20. Expanded the error message when iplsearch is called with an invalid key to
5786 suggest using net-iplsearch in a host list.
5788 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5789 ACLs are no longer actually imposed (and a message to that effect is
5792 22. If a "gecos" field in a passwd entry contained escaped characters, in
5793 particular, if it contained a \" sequence, Exim got it wrong when building
5794 a From: or a Sender: header from that name. A second bug also caused
5795 incorrect handling when an unquoted " was present following a character
5796 that needed quoting.
5798 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5799 was not being matched caselessly.
5801 24. Arranged for all hyphens in the exim.8 source to be escaped with
5804 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5805 properly. Recipient callout cache records were still being keyed to include
5806 the sender, even when use_sender was set false. This led to far more
5807 callouts that were necessary. The sender is no longer included in the key
5808 when use_sender is false.
5810 26. Added "control = submission" modifier to ACLs.
5812 27. Added the ${base62d: operator to decode base 62 numbers.
5814 28. dnsdb lookups can now access SRV records.
5816 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5817 the configuration file.
5819 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5820 (-v) mode. This makes the output for a verbose queue run more intelligible.
5822 31. Added a use_postmaster feature to recipient callouts.
5824 32. Added the $body_zerocount variable, containing the number of binary zero
5825 bytes in the message body.
5827 33. The time of last modification of the "new" subdirectory is now used as the
5828 "mailbox time last read" when there is a quota error for a maildir
5831 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5833 35. Added +ignore_unknown as a special item in host lists.
5835 36. Code for decoding IPv6 addresses in host lists is now included, even if
5836 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5837 address was recognized as an IP address, but was then not correctly decoded
5838 into binary, causing unexpected and incorrect effects when compared with
5845 1. Very minor rewording of debugging text in manualroute to say "list of
5846 hosts" instead of "hostlist".
5848 2. If verify=header_syntax was set, and a header line with an unqualified
5849 address (no domain) and a large number of spaces between the end of the
5850 name and the colon was received, the reception process suffered a buffer
5851 overflow, and (when I tested it) crashed. This was caused by some obsolete
5852 code that should have been removed. The fix is to remove it!
5854 3. When running in the test harness, delay a bit after writing a bounce
5855 message to get a bit more predictability in the log output.
5857 4. Added a call to search_tidyup() just before forking a reception process. In
5858 theory, someone could use a lookup in the expansion of smtp_accept_max_
5859 per_host which, without the tidyup, could leave open a database connection.
5861 5. Added the variables $recipient_data and $sender_data which get set from a
5862 lookup success in an ACL "recipients" or "senders" condition, or a router
5863 "senders" option, similar to $domain_data and $local_part_data.
5865 6. Moved the writing of debug_print from before to after the "senders" test
5868 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5869 problems for message scanning, either using a data ACL, or using
5870 local_scan() because the Received: header was not generated till after they
5871 were called (in order to set the time as the time of reception completion).
5872 I have revised the way this works. The header is now generated after the
5873 body is received, but before the ACL or local_scan() are called. After they
5874 are run, the timestamp in the header is updated.
5880 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5881 before starting a queue runner without re-exec. This happened only when
5882 deliver_drop_privilege was set or when the Exim user was set to root. The
5883 effect of the bug was that timeouts during subsequent deliveries caused
5884 crashes instead of being properly handled. The handler is now left at its
5885 default (and expected) setting.
5887 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5888 message, again when deliver_drop_privilege is set or Exim is run as root.
5889 The bug described in (1) was not present in this case, but the tidying up
5890 of the other signals was missing. I have made the two cases consistent.
5892 3. The ignore_target_hosts setting on a manualroute router was being ignored
5893 for hosts that were looked up using the /MX notation.
5895 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5898 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5899 operated on the sender address. After changing the $sender_address to <>
5900 for the sender address verify, Exim was re-instated it as the original
5901 (before rewriting) address, but remembering that it had rewritten it, so it
5902 wasn't rewriting it again. This bug also had the effect of breaking the
5903 sender address verification caching when the sender address was rewritten.
5905 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5906 This has been changed so that if the ip literal address matches
5907 ignore_target_hosts, the router declines.
5909 7. Added expansion conditions match_domain, match_address, and match_local_
5910 part (NOT match_host).
5912 8. The placeholder for the Received: header didn't have a length field set.
5914 9. Added code to Exim itself and to exim_lock to test for a specific race
5915 condition that could lead to file corruption when using MBX delivery. The
5916 issue is with the lockfile that is created in /tmp. If this file is removed
5917 after a process has opened it but before that process has acquired a lock,
5918 there is the potential for a second process to recreate the file and also
5919 acquire a lock. This could lead to two Exim processes writing to the file
5920 at the same time. The added code performs the same test as UW imapd; it
5921 checks after acquiring the lock that its file descriptor still refers to
5922 the same named file.
5924 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5925 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5928 11. Added the smtp_active_hostname option. If used, this will typically be made
5929 to depend on the incoming interface address. Because $interface_address is
5930 not set up until the daemon has forked a reception process, error responses
5931 that can happen earlier (such as "too many connections") no longer contain
5934 12. If an expansion in a condition on a "warn" statement fails because a lookup
5935 defers, the "warn" statement is abandoned, and the next ACL statement is
5936 processed. Previously this caused the whole ACL to be aborted.
5938 13. Added the iplsearch lookup type.
5940 14. Added ident_timeout as a log selector.
5942 15. Added tls_certificate_verified as a log selector.
5944 16. Added a global option tls_require_ciphers (compare the smtp transport
5945 option of the same name). This controls incoming TLS connections.
5947 17. I finally figured out how to make tls_require_ciphers do a similar thing
5948 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5949 before starting the TLS session.
5951 18. Tabs are now shown as \t in -bP output.
5953 19. If the log selector return_path_on_delivery was set, Exim crashed when
5954 bouncing a message because it had too many Received: header lines.
5956 20. If two routers both had headers_remove settings, and the first one included
5957 a superfluous trailing colon, the final name in the first list and the
5958 first name in the second list were incorrectly joined into one item (with a
5959 colon in the middle).
5965 1. Added -C and -D options to the exinext utility, mainly to make it easier
5966 to include in the automated testing, but these could be helpful when
5967 multiple configurations are in use.
5969 2. The exinext utility was not formatting the output nicely when there was
5970 an alternate port involved in the retry record key, nor when there was a
5971 message id as well (for retries that were specific to a specific message
5972 and a specific host). It was also confused by IPv6 addresses, because of
5973 the additional colons they contain. I have fixed the IPv4 problem, and
5974 patched it up to do a reasonable job for IPv6.
5976 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5977 delivery, the log line now contains "pipelined" if PIPELINING was used.
5979 4. An SMTP transport process used to panic and die if the bind() call to set
5980 an explicit outgoing interface failed. This has been changed; it is now
5981 treated in the same way as a connect() failure.
5983 5. A reference to $sender_host_name in the part of a conditional expansion
5984 that was being skipped was still causing a DNS lookup. This no longer
5987 6. The def: expansion condition was not recognizing references to header lines
5988 that used bh_ and bheader_.
5990 7. Added the _cache feature to named lists.
5992 8. The code for checking quota_filecount in the appendfile transport was
5993 allowing one more file than it should have been.
5995 9. For compatibility with Sendmail, the command line option
6003 and sets the incoming protocol and host name (for trusted callers). The
6004 host name and its colon can be omitted when only the protocol is to be set.
6005 Note the Exim already has two private options, -pd and -ps, that refer to
6006 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6007 "s", but I don't think that's a major issue.
6009 10. A number of refactoring changes to the code, none of which should affect
6012 (a) The number of logging options was getting close to filling up the
6013 32-bit word that was used as a bit map. I have split them into two classes:
6014 those that are passed in the argument to log_write(), and those that are
6015 only ever tested independently outside of that function. These are now in
6016 separate 32-bit words, so there is plenty of room for expansion again.
6017 There is no change in the user interface or the logging behaviour.
6019 (b) When building, for example, log lines, the code previously used a
6020 macro that called string_cat() twice, in order to add two strings. This is
6021 not really sufficiently general. Furthermore, there was one instance where
6022 it was actually wrong because one of the argument was used twice, and in
6023 one call a function was used. (As it happened, calling the function twice
6024 did not affect the overall behaviour.) The macro has been replaced by a
6025 function that can join an arbitrary number of extra strings onto a growing
6028 (c) The code for expansion conditions now uses a table and a binary chop
6029 instead of a serial search (which was left over from when there were very
6030 few conditions). Also, it now recognizes conditions like "pam" even when
6031 the relevant support is not compiled in: a suitably worded error message is
6032 given if an attempt is made to use such a condition.
6034 11. Added ${time_interval:xxxxx}.
6036 12. A bug was causing one of the ddress fields not to be passed back correctly
6037 from remote delivery subprocesses. The field in question was not being
6038 subsequently used, so this caused to problems in practice.
6040 13. Added new log selectors queue_time and deliver_time.
6042 14. Might have fixed a bug in maildirsizefile handling that threw up
6043 "unexpected character" debug warnings, and recalculated the data
6044 unnecessarily. In any case, I expanded the warning message to give more
6047 15. Added the message "Restricted characters in address" to the statements in
6048 the default ACL that block characters like @ and % in local parts.
6050 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6051 Three changes have been made:
6053 (a) There was a serious bug; a negative response to MAIL caused the whole
6054 recipient domain to be cached as invalid, thereby blocking all messages
6055 to all local parts at the same domain, from all senders. This bug has
6056 been fixed. The domain is no longer cached after a negative response to
6057 MAIL if the sender used is not empty.
6059 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6062 (c) A new callout option, "use_sender" has been added for people who want
6063 the modified behaviour.
6069 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6072 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6073 indeed breaks things for older releases.
6075 3. Added additional logging to the case where there is a problem reading data
6076 from a filter that is running in a subprocess using a pipe, in order to
6077 try to track down a specific problem.
6079 4. Testing facility fudge: when running in the test harness and attempting
6080 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6081 getting "No route to host". Convert this to a timeout.
6083 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6086 6. Some OS don't have socklen_t but use size_t instead. This affects the
6087 fifth argument of getsockopt() amongst other things. This is now
6088 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6089 can be set for individual OS. I have set it for SunOS5, OSF1, and
6090 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6091 some earlier ones do not.
6093 7. Change 4.30/15 was not doing the test caselessly.
6095 8. The standard form for an IPv6 address literal was being rejected by address
6096 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6097 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6098 this, as well as the form without the "IPv6" on the front (but only when
6099 address literals are enabled, of course).
6101 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6103 10. Exim crashed if a message with an empty sender address specified by -f
6104 encountered a router with an errors_to setting. This could be provoked only
6105 by a command such as
6109 where an empty string was supplied; "<>" did not hit this bug.
6111 11. Installed PCRE release 4.5.
6113 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6114 remained set. It is now erased.
6116 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6117 times from message ids (which are base 36 rather than the normal 62).
6119 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6120 were being counted as actual protocol errors, and logged if the log
6121 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6122 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6123 and DATA following a set of rejected RCPTs do not count as protocol errors.
6124 In other words, Exim assumes they were pipelined, though this may not
6125 actually be the case. Of course, in all cases the client gets an
6126 appropriate error code.
6128 15. If a lookup fails in an ACL condition, a message about the failure may
6129 be available; it is used if testing the ACL cannot continue, because most
6130 such messages specify what the cause of the deferral is. However, some
6131 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6132 that caused an old message to be retained and used if a later statement
6133 caused a defer, replacing the real cause of the deferral.
6135 16. If an IP address had so many PTR records that the DNS lookup buffer
6136 was not large enough to hold them, Exim could crash while trying to process
6137 the truncated data. It now detects and logs this case.
6139 17. Further to 4.21/58, another change has been made: if (and only if) the
6140 first line of a message (the first header line) ends with CRLF, a bare LF
6141 in a subsequent header line has a space inserted after it, so as not to
6142 terminate the header.
6144 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6145 unnecessarily, used atoi() instead of strtol(), and didn't check the
6146 termination when getting file sizes from file names by regex.
6148 19. Completely re-implemented the support for maildirsize files, in the light
6149 of a number of problems with the previous contributed implementation
6150 (4.30/29). In particular:
6152 . If the quota is zero, the maildirsize file is maintained, but no quota is
6155 . If the maildir directory does not exist, it is created before any attempt
6156 to write a maildirsize file.
6158 . The quota value in the file is just a cache; if the quota is changed in
6159 the transport, the new value overrides.
6161 . A regular expression is available for excluding directories from the
6164 20. The autoreply transport checks the characters in options that define the
6165 message's headers; it allows continued headers, but it was checking with
6166 isspace() after an embedded newline instead of explicitly looking for a
6169 21. If all the "regular" hosts to which an address was routed had passed their
6170 expiry times, and had not reached their retry times, the address was
6171 bounced, even if fallback hosts were defined. Now Exim should go on to try
6174 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6175 equivalent code in the SMTP transport. Some hosts send humungous responses
6176 to HELO/EHLO, more than 1024 it seems.
6178 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6179 gives compiler warnings in some environments. I've now done it "properly",
6182 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6183 (because of problems with the built-in one) was declared to return uschar *
6184 instead of char *, causing compiler failure.
6186 25. Fixed a file descriptor leak when processing alias/forward files.
6188 26. Fixed a minor format string issue in dbfn.c.
6190 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6192 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6193 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6196 29. When a sender address is verified, it is cached, to save repeating the test
6197 when there is more than one recipient in a message. However, when the
6198 verification involves a callout, it is possible for different callout
6199 options to be set for different recipients. It is too complicated to keep
6200 track of this in the cache, so now Exim always runs a verification when a
6201 callout is required, relying on the callout cache for the optimization.
6202 The overhead is duplication of the address routing, but this should not be
6205 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6206 to be verified with callout=postmaster, and the main callout worked but the
6207 postmaster check failed, the verification correctly failed. However, if a
6208 subsequent RCPT command asked for sender verification *without* the
6209 postmaster check, incorrect caching caused this verification also to fail,
6212 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6213 it was not caching the DNS options (qualify_single, search_parents) that
6214 were used when the lookup failed. A subsequent lookup with different
6215 options therefore always gave the same answer, though there were cases
6216 where it should not have. (Example: a "domains = !$mx_any" option on a
6217 dnslookup router: the "domains" option is always processed without any
6218 widening, but the router might have qualify_single set.) Now Exim uses the
6219 cached value only when the same options are set.
6221 32. Added John Jetmore's "exipick" utility to the distribution.
6223 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6224 than a timeout (e.g. a certificate is required, and is not provided), an
6225 Exim server now closes the connection immediately. Previously it waited for
6226 the client to close - but if the client is SSL, it seems that they each
6227 wait for each other, leading to a delay before one of them times out.
6229 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6230 maintained 0.8.x compatibility because I don't think many are using it, and
6231 it is clearly obsolete.
6233 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6236 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6237 client certificate was expired. A simple patch fixes this, though I don't
6238 understand the full logic of why the verify callback is called multiple
6241 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6242 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6243 which causes problems with some clients (such as the Certicom SSL Plus
6244 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6245 disables the coutermeasure allowing Eudora to connect."
6247 38. Exim was not checking that a write() to a log file succeeded. This could
6248 lead to Bad Things if a log got too big, in particular if it hit a file
6249 size limit. Exim now panics and dies if it cannot write to a log file, just
6250 as it does if it cannot open a log file.
6252 39. Modified OS/Makefile-Linux so that it now contains
6254 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6256 The two -D definitions ensure that Exim is compiled with large file
6257 support, which makes it possible to handle log files that are bigger than
6260 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6261 instance) a domain was checked against a named list that involved a lookup,
6262 causing $domain_data to be set, then another domain was checked against the
6263 same list, then the first domain was re-checked, the value of $domain_data
6264 after the final check could be wrong. In particular, if the second check
6265 failed, it could be set empty. This bug probably also applied to
6268 41. The strip_trailing_dot option was not being applied to the address given
6269 with the -f command-line option.
6271 42. The code for reading a message's header from the spool was incrementing
6272 $received_count, but never initializing it. This meant that the value was
6273 incorrect (doubled) while delivering a message in the same process in which
6274 it was received. In the most common configuration of Exim, this never
6275 happens - a fresh exec is done - but it can happen when
6276 deliver_drop_privilege is set.
6278 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6279 it now includes up to 150 characters of the unexpected data in the log
6282 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6283 and building data strings. The size of both of these buffers was 10 000
6284 bytes - far larger than anybody would *ever* want, thought I. Needless to
6285 say, somebody hit the limit. I have increased the maximum line length to
6286 20 000 and the maximum data length of concatenated lines to 100 000. I have
6287 also fixed two bugs, because there was no checking on these buffers. Tsk,
6288 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6289 buffer is too small.
6291 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6292 lsearch lookups. Now it does.
6294 46. When parsing a route_list item in a manualroute router, a fixed-length
6295 buffer was used for the list of hosts. I made this 1024 bytes long,
6296 thinking that nobody would ever have a list of hosts that long. Wrong.
6297 Somebody had a whole pile of complicated expansion conditions, and the
6298 string was silently truncated, leading to an expansion error. It turns out
6299 that it is easier to change to an unlimited length (owing to other changes
6300 that have happened since this code was originally written) than to build
6301 structure for giving a limitation error. The length of the item that
6302 expands into the list of hosts is now unlimited.
6304 47. The lsearch lookup could not handle data where the length of text line was
6305 more than 4095 characters. Such lines were truncated, leading to shortened
6306 data being returned. It should now handle lines of any length.
6308 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6309 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6312 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6313 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6314 no longer happens in comment lines. A list of replacements is now placed
6315 at the head of all of the source files, except those whose only change is
6316 to replace PERL_COMMAND in the very first #! line.
6318 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6319 messages on the queue, with a bottom-up merge sort, using code contributed
6320 by Michael Haardt. This should make operations like -bp somewhat faster on
6321 large queues. It won't affect queue runners, except when queue_run_in_order
6324 51. Installed eximstats 1.31 in the distribution.
6326 52. Added support for SRV lookups to the dnslookup router.
6328 53. If an ACL referred to $message_body or $message_body_end, the value was not
6329 reset for any messages that followed in the same SMTP session.
6331 54. The store-handling optimization for building very long strings was not
6332 differentiating between the different store pools. I don't think this
6333 actually made any difference in practice, but I've tidied it.
6335 55. While running the routers to verify a sender address, $sender_address
6336 was still set to the sender address. This is wrong, because when routing to
6337 send a bounce to the sender, it would be empty. Therefore, I have changed
6338 it so that, while verifying a sender address, $sender_address is set to <>.
6339 (There is no change to what happens when verifying a recipient address.)
6341 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6342 target A or AAAA records (if not already returned) without resetting the
6343 qualify_single or search_parents options of the DNS resolver. These are
6344 inappropriate in this case because the targets of MX and SRV records must
6345 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6346 target that, when qualified, matched something in the local domain. These
6347 two options are now turned off when doing these lookups.
6349 57. It seems that at least some releases of Reiserfs (which does not have the
6350 concept of a fixed number of inodes) returns zero and not -1 for the
6351 number of available inodes. This interacted badly with check_spool_inodes,
6352 which assumed that -1 was the "no such thing" setting. What I have done is
6353 to check that the total number of inodes is greater than zero before doing
6354 the test of how many are available.
6356 58. When a "warn" ACL statement has a log_message modifier, the message is
6357 remembered, and not repeated. This is to avoid a lot of repetition when a
6358 message has many recipients that cause the same warning to be written.
6359 However, Exim was preserving the list of already written lines for an
6360 entire SMTP session, which doesn't seem right. The memory is now reset if a
6361 new message is started.
6363 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6364 debugging output unless log_rewrite was also set.
6366 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6367 of (int)(handle) when we know that handle contains (void *)(-1).
6369 61. The Exim daemon panic-logs an error return when it closes the incoming
6370 connection. However "connection reset by peer" seems to be common, and
6371 isn't really an error worthy of noting specially, so that particular error
6374 62. When Exim is trying to find all the local interfaces, it used to panic and
6375 die if the ioctl to get the interface flags failed. However, it seems that
6376 on at least one OS (Solaris 9) it is possible to have an interface that is
6377 included in the list of interfaces, but for which you get a failure error
6378 for this call. This happens when the interface is not "plumbed" into a
6379 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6380 failure of the "get flags" call assumes that the interface is down.
6382 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6383 makes life easier for people who are doing arithmetic on fields extracted
6384 from dates, where you often get leading zeros that should not be
6385 interpreted as octal.
6387 64. Added qualify_domain to the redirect router, to override the global
6390 65. If a pathologically long header line contained very many addresses (the
6391 report of this problem mentioned 10 000) and each of them was rewritten,
6392 Exim could use up a very large amount of memory. (It kept on making new
6393 copies of the header line as it rewrote, and never released the old ones.)
6394 At the expense of a bit more processing, the header rewriting function has
6395 been changed so that it no longer eats memory in this way.
6397 66. The generation of the Received: header has been moved from the time that a
6398 message starts to be received, to the time that it finishes. The timestamp
6399 in the Received: header should now be very close to that of the <= log
6400 line. There are two side-effects of this change:
6402 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6403 logged header lines no longer include the local Received: line, because
6404 it has not yet been created. The same applies to a copy of the message
6405 that is returned to a non-SMTP sender when a message is rejected.
6407 (b) When a filter file is tested using -bf, no additional Received: header
6408 is added to the test message. After some thought, I decided that this
6411 This change does not affect the value of $received_for. It is still set
6412 after address rewriting, but before local_scan() is called.
6414 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6416 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6417 gave an unhelpful panic error message, and a defer error. I have managed to
6418 change this behaviour so that it now rejects any supplied certificate,
6419 which seems right, as the list of acceptable certificates is empty.
6421 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6422 gave an unhelpful defer error. I have not managed to make this reject any
6423 supplied certificates, but the error message it gives is "no certificate
6424 supplied", which is not helpful.
6426 70. exigrep's output now also includes lines that are not associated with any
6427 message, but which match the given pattern. Implemented by a patch from
6428 Martin Sluka, which also tidied up the Perl a bit.
6430 71. Recipient callout verification, like sender verification, was using <> in
6431 the MAIL FROM command. This isn't really the right thing, since the actual
6432 sender may affect whether the remote host accepts the recipient or not. I
6433 have changed it to use the actual sender in the callout; this means that
6434 the cache record is now keyed on a recipient/sender pair, not just the
6435 recipient address. There doesn't seem to be a real danger of callout loops,
6436 since a callout by the remote host to check the sender would use <>.
6437 [SEE ABOVE: changed after hitting problems.]
6439 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6440 temporary errors. However, in the case of such a code being given after
6441 the end of a data transmission (i.e. after ".") Exim was failing to write
6442 a retry record for the message. (Yes, there was some broken host that was
6443 actually sending 8xx at this point.)
6445 73. An unknown lookup type in a host list could cause Exim to panic-die when
6446 the list was checked. (An example that provoked this was putting <; in the
6447 middle of a list instead of at the start.) If this happened during a DATA
6448 ACL check, a -D file could be left lying around. This kind of configuration
6449 error no longer causes Exim to die; instead it causes a defer error. The
6450 incident is still logged to the main and panic logs.
6452 74. Buglet left over from Exim 3 conversion. The message "too many messages
6453 in one connection" was written to the rejectlog but not the mainlog, except
6454 when address rewriting (yes!) was being logged.
6456 75. Added write_rejectlog option.
6458 76. When a system filter was run not as root (that is, when system_filter_user
6459 was set), the values of the $n variables were not being returned to the
6460 main process; thus, they were not subsequently available in the $sn
6463 77. Added +return_path_on_delivery log selector.
6465 78. A connection timeout was being treated differently from recipients deferred
6466 when testing hosts_max_try with a message that was older than the host's
6467 retry timeout. (The host should not be counted, thus allowing all hosts to
6468 be tried at least once before bouncing.) This may have been the cause of an
6469 occasionally reported bug whereby a message would remain on the queue
6470 longer than the retry timeout, but would be bounced if a delivery was
6471 forced. I say "may" because I never totally pinned down the problem;
6472 setting up timeout/retry tests is difficult. See also the next item.
6474 79. The ultimate address timeout was not being applied to errors that involved
6475 a combination of host plus message (for example, a timeout on a MAIL
6476 command). When an address resolved to a number of possible hosts, and they
6477 were not all tried for each delivery (e.g. because of hosts_max_try), a
6478 message could remain on the queue longer than the retry timeout.
6480 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6483 81. Fixed an obscure SMTP outgoing bug which required at least the following
6484 conditions: (a) there was another message waiting for the same server;
6485 (b) the server returned 5xx to all RCPT commands in the first message so
6486 that the message was not completed; (c) the server dropped the connection
6487 or gave a negative response to the RSET that Exim sends to abort the
6488 transaction. The observed case was a dropped connection after DATA that had
6489 been sent in pipelining mode. That is, the server had advertised PIPELINING
6490 but was not implementing it correctly. The effect of the bug was incorrect
6491 behaviour, such as trying another host, and this could lead to a crash.
6497 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6498 and daemon.c were passed as pointers to ints; they should have been
6499 pointers to socklen_t variables (which are typically unsigned ints).
6501 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6504 3. Fixed a really odd bug that affected only the testing scheme; patching a
6505 certain fixed string in the binary changed the value of another string that
6506 happened to be identical to the end of the original first string.
6508 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6509 name", it returns that address as the IP address. On some operating
6510 systems (e.g. Solaris), it also passes back the IP address string as the
6511 "host name". However, on others (e.g. Linux), it passes back an empty
6512 string. Exim wasn't checking for this, and was changing the host name to an
6513 empty string, assuming it had been canonicalized.
6515 5. Although rare, it is permitted to have more than one PTR record for a given
6516 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6517 all the names associated with an address, because they do in Solaris.
6518 However, it seems that they do not in Linux for data that comes from the
6519 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6520 I found this out when I moved to a new Linux workstation and tried to run
6521 the Exim test suite.
6523 To get round this problem I have changed the code so that it now does its
6524 own call to the DNS to look up PTR records when searching for a host name.
6525 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6526 addresses that are only in /etc/hosts are still found.
6528 This behaviour is, however, controlled by an option called host_lookup_
6529 order, which defaults to "bydns:byaddr". If people want to use the other
6530 order, or indeed, just use one or the other means of lookup, they can
6531 specify it in this variable.
6533 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6534 some operating systems, this comes back from gethostbyaddr() as an empty
6535 string, and this is what Exim used to test for. However, it seems that in
6536 other systems, "." is yielded. Exim now tests for this case too.
6538 7. The values of check_spool_space and check_log_space are now held internally
6539 as a number of kilobytes instead of an absolute number of bytes. If a
6540 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6541 kilobyte. This means that much larger values can be stored.
6543 8. Exim monitor: an attempt to get the action menu when not actually pointing
6544 at a message produces an empty menu entitled "No message selected". This
6545 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6546 no entries in it ("Shell widget menu has zero width and/or height"). So I
6547 have added a single, blank menu entry in this case.
6549 9. Added ${quote_local_part.
6551 10. MIME decoding is now applied to the contents of Subject: header lines when
6554 11. Now that a reference to $sender_host_address automatically causes a reverse
6555 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6556 host lookup before query-style lookups in lists that might use this
6557 variable. This has therefore been abolished, and the "net-" prefix is no
6558 longer necessary for query-style lookups.
6560 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6561 have been a typo for LFLAGS, so it has been changed.
6563 13. The install script calls Exim with "-C /dev/null" in order to find the
6564 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6565 to be output. However, since Exim outputs its version number before the
6566 error, it didn't break the script. It just looked ugly. I fixed this by
6567 always allowing "-C /dev/null" if the caller is root.
6569 14. Ignore overlarge ACL variable number when reading spool file - insurance
6570 against a later release with more variables having written the file.
6572 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6573 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6574 this, as well as the form without the "IPv6" on the front.
6576 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6577 OS/Makefile-Darwin file.
6579 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6580 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6582 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6583 variable. However, some OS use ints for inodes. I've added cast to long int
6584 to get rid of the compiler warning.
6586 19. I had forgotten to lock out "/../" in configuration file names when
6587 ALT_CONFIG_PREFIX was set.
6589 20. Routers used for verification do not need to specify transports. However,
6590 if such a router generated a host list, and callout was configured, Exim
6591 crashed, because it could not find a port number from the (non-existent)
6592 transport. It now assumes port 25 in this circumstance.
6594 21. Added the -t option to exigrep.
6596 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6597 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6598 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6599 them is set, LOOKUP_LSEARCH is forced.
6601 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6603 24. Added sender and host information to the "rejected by local_scan()" log
6604 line; previously there was no indication of these.
6606 25. Added .include_if_exists.
6608 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6609 Linux. It turns out that not all file systems support this. Apparently some
6610 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6611 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6612 not supported on the file descriptor, is now ignored when Exim is trying to
6613 sync a directory. This applies only to Linux.
6615 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6617 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6618 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6621 29. Added support for maildirsize files from supplied patch (modified a bit).
6623 30. The use of :fail: followed by an empty string could lead Exim to respond to
6624 sender verification failures with (e.g.):
6626 550 Verification failed for <xxx>
6627 550 Sender verify failed
6629 where the first response line was missing the '-' that indicates it is not
6630 the final line of the response.
6632 31. The loop for finding the name of the user that called Exim had a hardwired
6633 limit of 10; it now uses the value of finduser_retries, which is used for
6634 all other user lookups.
6636 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6639 33. Exim was neglecting to zero errno before one call of strtol() when
6640 expanding a string and expecting an integer value. On some systems this
6641 resulted in spurious "integer overflow" errors. Also, it was casting the
6642 result into an int without checking.
6644 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6645 did not work. The code looks as if it has *never* worked, though it appears
6646 to have been documented since at least release 1.62. I have made it work.
6648 35. The "timeout_DNS" error in retry rules, also documented since at least
6649 1.62, also never worked. As it isn't clear exactly what this means, and
6650 clearly it isn't a major issue, I have abolished the feature by treating it
6651 as "timeout", and writing a warning to the main and panic logs.
6653 36. The display of retry rules for -brt wasn't always showing the error code
6656 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6657 timeout_connect_A, timeout_connect_MX.
6659 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6660 to the empty sender.
6662 39. The daemon was not analysing the content of -oX till after it had closed
6663 stderr and disconnected from the controlling terminal. This meant that any
6664 syntax errors were only noted on the panic log, and the return code from
6665 the command was 0. By re-arranging the code a little, I've made the
6666 decoding happen first, so such errors now appear on stderr, and the return
6667 code is 1. However, the actual setting up of the sockets still happens in
6668 the disconnected process, so errors there are still only recorded on the
6671 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6672 connections (as happens on some IP stacks) was logged at start up time as
6673 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6674 it from "IPv6 and IPv4", which means that two separate sockets are being
6677 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6678 says whether AF_INET or AF_INET6 was passed as an argument.
6680 42. Exiwhat output was messed up when time zones were included in log
6683 43. Exiwhat now gives more information about the daemon's listening ports,
6684 and whether -tls-on-connect was used.
6686 44. The "port" option of the smtp transport is now expanded.
6688 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6689 silently ignored. Now an error message is written to the main and panic
6692 46. There's a new ACL modifier called "logwrite" which writes to a log file
6693 as soon as it is encountered.
6695 47. Added $local_user_uid and $local_user_gid at routing time.
6697 48. Exim crashed when trying to verify a sender address that was being
6700 49. Exim was recognizing only a space character after ".include". It now also
6701 recognizes a tab character.
6703 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6704 extracting the relevant information from the specification. The man page no
6705 longer contains scrambled data for the -d option, and I've added a section
6706 at the front about calling Exim under different names.
6708 51. Added "extra_headers" argument to the "mail" command in filter files.
6710 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6713 53. Installed eximstats 1.29.
6715 54. Added transport_filter_timeout as a generic transport option.
6717 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6718 Cc: header lines. This was required by RFC 822, but it not required by RFC
6721 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6722 incoming messages that did not have them. Now it does so only if the
6723 message originates locally, that is, if there is no associated remote host
6724 address. When Resent- header lines are present, this applies to the Resent-
6725 lines rather than the non-Resent- lines.
6727 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6728 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6730 58. Messages for configuration errors now include the name of the main
6731 configuration file - useful now that there may be more than one file in a
6732 list (.included file names were always shown).
6734 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6735 for those rare installations that do not start the daemon as root or run it
6736 setuid root. I've cut out the call to initgroups() if the daemon is not
6739 60. The Exim user and group can now be bound into the binary as text strings
6740 that are looked up at the start of Exim's processing.
6742 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6744 62. Added $mailstore_basename variable.
6746 63. Installed patch to sieve.c from Michael Haardt.
6748 64. When Exim failed to open the panic log after failing to open the main log,
6749 the original message it was trying to log was written to stderr and debug
6750 output, but if they were not available (the usual case in production), it
6751 was lost. Now it is written to syslog before the two lines that record the
6752 failures to open the logs.
6754 65. Users' Exim filters run in subprocesses under the user's uid. It is
6755 possible for a "deliver" command or an alias in a "personal" command to
6756 provoke an address rewrite. If logging of address rewriting is configured,
6757 this fails because the process is not running as root or exim. There may be
6758 a better way of dealing with this, but for the moment (because 4.30 needs
6759 to be released), I have disabled address rewrite logging when running a
6760 filter in a non-root, non-exim process.
6766 1. The buildconfig auxiliary program wasn't quoting the value set for
6767 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6768 not defined. This bug was masked in 4.22 by the effect that was fixed in
6771 2. Some messages that were rejected after a message id was allocated were
6772 shown as "incomplete" by exigrep. It no longer does this for messages that
6773 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6775 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6776 have allow_domain_literals set, the ID did not get logged in the <= line.
6777 Domain literals are now always recognized in Message-ID: header lines.
6779 4. The first argument for a ${extract expansion item is the key name or field
6780 number. Leading and trailing spaces in this item were not being ignored,
6781 causing some misleading effects.
6783 5. When deliver_drop_privilege was set, single queue runner processes started
6784 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6785 same command in the process it spins off) were not dropping privilege.
6787 6. When the daemon running as "exim" started a queue runner, it always
6788 re-executed Exim in the spun-off process. This is a waste of effort when
6789 deliver_drop_privilege is set. The new process now just calls the
6790 queue-runner function directly.
6796 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6799 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6800 ignored. Though the use of -oP was forcing the writing of a pid file, it
6801 was always written to the default place.
6803 3. If the message "no IP address found for host xxxx" is generated during
6804 incoming verification, it is now followed by identification of the incoming
6805 connection (so you can more easily find what provoked it).
6807 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6809 5. Added some features to "harden" Exim a bit more against certain attacks:
6811 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6812 be put in Local/Makefile. This is like the never_users runtime option,
6813 but it cannot be overridden. The default setting is "root".
6815 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6816 prefix string with which any file named in a -C command line option
6819 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6820 is retained for -C and -D only if the caller of Exim is root. Without
6821 it, the exim user may also use -C and -D and retain privilege.
6823 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6824 command line option is disabled.
6826 6. Macro names set by the -D option must start with an upper case letter, just
6827 like macro names defined in the configuration file.
6829 7. Added "dereference=" facility to LDAP.
6831 8. Two instances of the typo "uknown" in the source files are fixed.
6833 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6834 the Configure-Makefile script screwed up while processing it.
6836 10. Incorporated PCRE 4.4.
6838 11. The SMTP synchronization check was not operating right at the start of an
6839 SMTP session. For example, it could not catch a HELO sent before the client
6840 waited for the greeting. There is now a check for outstanding input at the
6841 point when the greeting is written. Because of the duplex, asynchronous
6842 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6843 way, but not yet received, when the check is performed.
6845 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6846 on TCP/IP sockets, because this apparently causes some broken clients to
6849 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6850 unchanged) from the Cygwin maintainer.
6852 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6853 was supported instead of testing for mbx. Effectively a typo.
6855 15. The spa authenticator server code was not checking that the input it
6856 received was valid base64.
6858 16. The debug output line for the "set" modifier in ACLs was not showing the
6859 name of the variable that was being set.
6861 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6863 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6864 lookup on its own. Something else was needed to trigger the lookup. For
6865 example, a match in host_lookup or the need for a host name in a host list.
6866 Now, if $sender_host_name is referenced and the host name has not yet been
6867 looked up, a lookup is performed. If the lookup fails, the variable remains
6868 empty, and $host_lookup_failed is set to "1".
6870 19. Added "eqi" as a case-independent comparison operator.
6872 20. The saslauthd authentication condition could segfault if neither service
6873 nor realm was specified.
6875 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6876 error message that was logged was misleading, and incoming SMTP
6877 connections were dropped. The message is now more accurate, and temporary
6878 errors are given to SMTP connections.
6880 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6881 (except RSET) with a 421 error, until QUIT is received. However, it was
6882 failing to send a response to QUIT.
6884 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6885 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6886 ACL is now run after the helo_try_verify_hosts code.
6888 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6889 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6890 case-independent, so other case variants are also recognized). Apparently
6891 some systems use these upper case variants.
6893 25. If more than two messages were waiting for the same host, and a transport
6894 filter was specified for the transport, Exim sent two messages over the
6895 same TCP/IP connection, and then failed with "socket operation on non-
6896 socket" when it tried to send the third.
6898 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6900 27. The extern definition of crypt16() in expand.c was not being excluded when
6901 the OS had its own crypt16() function.
6903 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6904 as a preferred synonym for return_size_limit, both as an option and as an
6907 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6909 30. Changed the default configuration ACL to relax the local part checking rule
6910 for addresses that are not in any local domains. For these addresses,
6911 slashes and pipe symbols are allowed within local parts, but the sequence
6912 /../ is explicitly forbidden.
6914 31. SPA server authentication was not clearing the challenge buffer before
6917 32. log_message in a "warn" ACL statement was writing to the reject log as
6918 well as to the main log, which contradicts the documentation and doesn't
6919 seem right (because no rejection is happening). So I have stopped it.
6921 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6922 However, I am unable to do any testing of this.
6924 34. Fixed an infelicity in the appendfile transport. When checking directories
6925 for a mailbox, to see if any needed to be created, it was accidentally
6926 using path names with one or more superfluous leading slashes; tracing
6927 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6929 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6930 contents are added to the log line that is written for every discarded
6931 recipient. (Previously a log_message setting was ignored.)
6933 36. The ${quote: operator now quotes the string if it is empty.
6935 37. The install script runs exim in order to find its version number. If for
6936 some reason other than non-existence or emptiness, which it checks, it
6937 could not run './exim', it was installing it with an empty version number,
6938 i.e. as "exim-". This error state is now caught, and the installation is
6941 38. An argument was missing from the function that creates an error message
6942 when Exim fails to connect to the socket for saslauthd authentication.
6943 This could cause Exim to crash, or give a corrupted message.
6945 39. Added isip, isip4, and isip6 to ${if conditions.
6947 40. The ACL variables $acl_xx are now saved with the message, and can be
6948 accessed later in routers, transports, and filters.
6950 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6951 strings in the file are not string-expanded.
6953 42. If a MAIL command specified a SIZE value that was too large to fit into an
6954 int variable, the check against message_size_limit failed. Such values are
6955 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6956 day this will have to be increased, but I don't think I want to be around
6957 when emails are that large.
6964 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6965 iconv() is not standard in FreeBSD.
6967 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6968 IPv6 enabled. The observed symptom was a segmentation fault on return from
6969 the function os_common_find_running_interfaces() in src/os.c.
6971 3. In the check_special_case() function in daemon.c I had used "errno" as an
6972 argument name, which causes warnings on some systems. This was basically a
6973 typo, since it was named "eno" in the comments!
6975 4. The code that waits for the clock to tick (at a resolution of some fraction
6976 of a second) so as to ensure message-id uniqueness was always waiting for
6977 at least one whole tick, when it could have waited for less. [This is
6978 almost certainly not relevant at current processor speeds, where it is
6979 unlikely to ever wait at all. But we try to future-proof.]
6981 5. The function that sleeps for a time interval that includes fractions of a
6982 second contained a race. It did not block SIGALRM between setting the
6983 timer, and suspending (a couple of lines later). If the interval was short
6984 and the sigsuspend() was delayed until after it had expired, the suspension
6985 never ended. On busy systems this could lead to processes getting stuck for
6988 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6989 process, before it forks any delivery processes. The open lookup caching
6990 mechanism meant that the open file or database connection was passed into
6991 the delivery process. The problem was that delivery processes always tidy
6992 up cached lookup data. This could cause a problem for the next delivery
6993 process started by the queue runner, because the external queue runner
6994 process does not know about the closure. So the next delivery process
6995 still has data in the lookup cache. In the case of a file lookup, there was
6996 no problem because closing a file descriptor in a subprocess doesn't affect
6997 the parent. However, if the lookup was caching a connection to a database,
6998 the connection was closed, and the second delivery process was likely to
6999 see errors such as "PGSQL: query failed: server closed the connection
7000 unexpectedly". The problem has been fixed by closing all cached lookups
7001 in a queue runner before running a delivery process.
7003 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7004 seem to have the "const" qualifier which it has on other OS. I've
7007 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7008 *of the same type* (client or server) with the same public name that an
7009 error should be diagnosed.
7011 9. When Exim looked up a host name for an IP address, but failed to find the
7012 original IP address when looking up the host name (a safety check), it
7013 output the message "<ip address> does not match any IP for NULL", which was
7014 confusing, to say the least. The bug was that the host name should have
7015 appeared instead of "NULL".
7017 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7018 user that is built into the binary, and the -C or -D options is used, root
7019 privilege is dropped before the configuration file is read. In addition,
7020 logging is switched to stderr instead of the normal log files. If the
7021 configuration then re-defines the Exim user, the unprivileged environment
7022 is probably not what is expected, so Exim logs a panic warning message (but
7025 However, if deliver_drop_privilege is set, the unprivileged state may well
7026 be exactly what is intended, so the warning has been cut out in that case,
7027 and Exim is allowed to try to write to its normal log files.
7033 1. smtp_return_error_details was not giving details for temporary sender
7034 or receiver verification errors.
7036 2. Diagnose a configuration error if two authenticators have the same public
7039 3. Exim used not to create the message log file for a message until the first
7040 delivery attempt. This could be confusing when incoming messages were held
7041 for policy or load reasons. The message log file is now created at the time
7042 the message is received, and an initial "Received" line is written to it.
7044 4. The automatically generated man page for command line options had a minor
7045 bug that caused no ill effects; however, a more serious problem was that
7046 the procedure for building the man page automatically didn't always
7047 operate. Consequently, release 4.20 contains an out-of-date version. This
7048 shouldn't happen again.
7050 5. When building Exim with embedded Perl support, the script that builds the
7051 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7052 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7054 6. The freeze_tell option was not being used for messages that were frozen on
7055 arrival, either by an ACL or by local_scan().
7057 7. Added the smtp_incomplete_transaction log selector.
7059 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7060 was accepting AUTH without a new EHLO.
7062 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7063 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7066 10. Logging of TCP/IP connections (when configured) now happens in the main
7067 daemon process instead of the child process, so that the TCP/IP connection
7068 count is more accurate (but it can never be perfect).
7070 11. The use of "drop" in a nested ACL was not being handled correctly in the
7071 outer ACL. Now, if condition failure induced by the nested "drop" causes
7072 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7073 or "require"), the connection is dropped.
7075 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7076 that yield "discard" can now be used with an "accept" or a "discard" verb,
7077 but an error is generated for any others (because I can't see a useful way
7078 to define what should happen).
7080 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7081 are now processed in the same way as lines in the Exim configuration file.
7082 In particular, continuation lines are supported.
7084 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7086 15. Added -ti meaning -t -i.
7088 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7089 domains, and warn by logging if others are found.
7091 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7092 returned from GIFCONF when Exim is trying to find the list of interfaces on
7093 a host. The code in os.c has been modified to copy each ifreq to an aligned
7094 structure in all cases.
7096 Also, in some cases, the returned ifreq's were being copied to a 'struct
7097 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7098 means the last couple of bytes of an IPv6 address could be chopped if the
7099 ifreq contained only a normal sockaddr (14 bytes storage).
7101 18. Named domain lists were not supported in the hosts_treat_as_local option.
7102 An entry such as +xxxx was not recognized, and was treated as a literal
7105 19. Ensure that header lines added by a DATA ACL are included in the reject log
7106 if the ACL subsequently rejects the message.
7108 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7109 MD5 (which is deprecated).
7111 21. When testing a filter file using -bf, Exim was writing a message when it
7112 took the sender from a "From " line in the message, but it was not doing so
7113 when it took $return_path from a Return-Path: header line. It now does.
7115 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7116 with a valid header line field name (a series of printing characters
7117 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7119 23. Changed "disc" in the source to "disk" to conform to the documentation and
7120 the book and for uniformity.
7122 24. Ignore Sendmail's -Ooption=value command line item.
7124 25. When execve() failed while trying to run a command in a pipe transport,
7125 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7126 could be confused with a return value of 69 from the command itself. This
7127 has been changed to 127, the value the shell returns if it is asked to run
7128 a non-existent command. The wording for the related log line suggests a
7129 non-existent command as the problem.
7131 26. If received_header_text expands to an empty string, do not add a Received:
7132 header line to the message. (Well, it adds a token one on the spool, but
7133 marks it "old" so that it doesn't get used or transmitted.)
7135 27. Installed eximstats 1.28 (addition of -nt option).
7137 28. There was no check for failure on the call to getsockname() in the daemon
7138 code. This can fail if there is a shortage of resources on the system, with
7139 ENOMEM, for example. A temporary error is now given on failure.
7141 29. Contrary to the C standard, it seems that in some environments, the
7142 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7143 program. Exim now does this explicitly; it affects the formatting of
7144 timestamps using strftime().
7146 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7147 complaints. I've now initialized all the variables, to avoid this.
7149 32. Header lines added by a system filter were not being "seen" during
7150 transport-time rewrites.
7152 33. The info_callback() function passed to OpenSSL is set up with type void
7153 (*)(SSL *, int, int), as described somewhere. However, when calling the
7154 function (actually a macro) that sets it up, the type void(*)() is
7155 expected. I've put in a cast to prevent warnings from picky compilers.
7157 34. If a DNS black list lookup found a CNAME record, but there were no A
7158 records associated with the domain it pointed at, Exim crashed.
7160 35. If a DNS black list lookup returned more than one A record, Exim ignored
7161 all but the first. It now scans all returned addresses if a particular IP
7162 value is being sought. In this situation, the contents of the
7163 $dnslist_value variable are a list of all the addresses, separated by a
7166 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7167 to accept a host name and all its aliases if the forward lookup for any of
7168 them yielded the IP address of the incoming connection. Now it accepts only
7169 those names whose forward lookup yields the correct IP address. Any other
7170 names are discarded. This closes a loophole whereby a rogue DNS
7171 administrator could create reverse DNS records to break through a
7172 wildcarded host restriction in an ACL.
7174 37. If a user filter or a system filter that ran in a subprocess used any of
7175 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7176 the wrong values were passed to the pipe command ($thisaddress had the
7177 value of $0, $0 had the value of $1, etc). This bug was introduced by
7178 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7180 38. Improved the line breaking for long SMTP error messages from ACLs.
7181 Previously, if there was no break point between 40 and 75 characters, Exim
7182 left the rest of the message alone. Two changes have been made: (a) I've
7183 reduced the minimum length to 35 characters; (b) if it can't find a break
7184 point between 35 and 75 characters, it looks ahead and uses the first one
7185 that it finds. This may give the occasional overlong line, but at least the
7186 remaining text gets split now.
7188 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7189 file descriptors might be low, and that setting 1000 would always raise it.
7190 It turns out that in some environments, the limit is already over 1000 and
7191 that lowering it causes trouble. So now Exim takes care not to decrease it.
7193 40. When delivering a message, the value of $return_path is set to $sender_
7194 address at the start of routing (routers may change the value). By an
7195 oversight, this default was not being set up when an address was tested by
7196 -bt or -bv, which affected the outcome if any router or filter referred to
7199 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7200 log only when "message" or "log_message" are set. However, if one of the
7201 conditions was an address verification, or a call to a nested ACL, the
7202 messages generated by the underlying test were being passed through. This
7203 no longer happens. The underlying message is available in $acl_verify_
7204 message for both "message" and "log_message" expansions, so it can be
7205 passed through if needed.
7207 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7208 new expansion $bh_ to give the encoded byte string without charset
7209 translation. Translation happens only if iconv() is available; HAVE_ICONV
7210 indicates this at build time. HEADERS_CHARSET gives the charset to
7211 translate to; headers_charset can change it in the configuration, and
7212 "headers charset" can change it in an individual filter file.
7214 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7215 that creates RFC 2047 encoded "words" labels them as that charset instead
7216 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7217 expansion operator; (ii) when Exim creates a From: line for a local
7218 message; (iii) when a header line is rewritten to include a "phrase" part.
7220 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7221 buggy, causing it to skip the first lines of messages whose message ID
7222 ended in 'D'. This would not have bitten before Exim release 4.14, because
7223 message IDs were unlikely to end in 'D' before then. The effect was to have
7224 incorrect size information for certain domains.
7226 45. #include "config.h" was missing at the start of the crypt16.c module. This
7227 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7230 46. If there was a timeout during a "random" callout check, Exim treated it as
7231 a failure of the random address, and carried on sending RSET and the real
7232 address. If the delay was just some slowness somewhere, the response to the
7233 original RCPT would be taken as a response to RSET and so on, causing
7234 mayhem of various kinds.
7236 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7237 when I implemented it. It didn't allow for the fact that some option values
7238 may legitimately be negative (e.g. size_addition), and it didn't even do
7239 the right test for positive values.
7241 48. Domain names in DNS records are case-independent. Exim always looks them up
7242 in lower case. Some resolvers return domain names in exactly the case they
7243 appear in the zone file, that is, they may contain uppercase letters. Not
7244 all resolvers do this - some return always lower case. Exim was treating a
7245 change of case by a resolver as a change of domain, similar to a widening
7246 of a domain abbreviation. This triggered its re-routing code and so it was
7247 trying to route what was effectively the same domain again. This normally
7248 caused routing to fail (because the router wouldn't handle the domain
7249 twice). Now Exim checks for this case specially, and just changes the
7250 casing of the domain that it ultimately uses when it transmits the message
7253 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7256 50. If a filter generated a file delivery with a non-absolute name (possible if
7257 no home directory exists for the router), the forbid_file option was not
7260 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7261 the existing equality matching.
7263 52. Exim was using ints instead of ino_t variables in some places where it was
7264 dealing with inode numbers.
7266 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7267 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7268 called TMPDIR, and if it finds it is different, it changes its value.
7270 54. The smtp_printf() function is now made available to local_scan() so
7271 additional output lines can be written before returning. There is also an
7272 smtp_fflush() function to enable the detection of a dropped connection.
7273 The variables smtp_input and smtp_batched_input are exported to
7276 55. Changed the default runtime configuration: the message "Unknown user"
7277 has been removed from the ACL, and instead placed on the localuser router,
7278 using the cannot_route_message feature. This means that any verification
7279 failures that generate their own messages won't get overridden. Similarly,
7280 the "Unrouteable address" message that was in the ACL for unverifiable
7281 relay addresses has also been removed.
7283 56. Added hosts_avoid_esmtp to the smtp transport.
7285 57. The exicyclog script was not checking for the esoteric option
7286 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7287 will work only if exicyclog is run under the appropriate euid.
7289 58. Following a discussion on the list, the rules by which Exim recognises line
7290 endings on incoming messages have been changed. The -dropcr and drop_cr
7291 options are now no-ops, retained only for backwards compatibility. The
7292 following line terminators are recognized: LF CRLF CR. However, special
7293 processing applies to CR:
7295 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7296 nor a local message in the state where . is a terminator.
7298 (ii) If a bare CR is encountered in a header line, an extra space is added
7299 after the line terminator so as not to end the header. The reasoning
7300 behind this is that bare CRs in header lines are most likely either
7301 to be mistakes, or people trying to play silly games.
7303 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7304 was being incorrectly given as 18 bytes larger than it should have been.
7305 This is a VOB (very old bug).
7307 60. This may never have affected anything current, but just in case it has:
7308 When the local host is found other than at the start of a list of hosts,
7309 the local host, those with the same MX, and any that follow, are discarded.
7310 When the list in question was part of a longer list of hosts, the following
7311 hosts (not currently being processed) were also being discarded. This no
7312 longer happens. I'm not sure if this situation could ever has previously
7315 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7318 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7319 header. This is something that is recommended in a new Internet Draft, and
7320 is something that is documented as being done by Sendmail. There are two
7321 possible values. For messages generated by the autoreply transport, Exim
7324 Auto-Submitted: auto-replied
7326 whereas for all other generated messages (e.g. bounces) it adds
7328 Auto-Submitted: auto-generated
7330 63. The "personal" condition in filters now includes a test for the
7331 Auto-Submitted: header. If it contains the string "auto-" the message it
7332 not considered personal.
7334 64. Added rcpt_include_affixes as a generic transport option.
7336 65. Added queue_only_override (default true).
7338 66. Added the syslog_duplication option.
7340 67. If what should have been the first header line of a message consisted of
7341 a space followed by a colon, Exim was mis-interpreting it as a header line.
7342 It isn't of course - it is syntactically invalid and should therefore be
7343 treated as the start of the message body. The misbehaviour could have
7344 caused a number of strange effects, including loss of data in subsequent
7345 header lines, and spool format errors.
7347 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7348 client host had authenticated. This control can now be exercised by an ACL
7349 for more flexibility.
7351 69. By default, callouts do not happen when testing with -bh. There is now a
7352 variant, -bhc, which does actually run the callout code, including
7353 consulting and updating the callout cache.
7355 70. Added support for saslauthd authentication, courtesy of Alexander
7358 71. If statvfs() failed on the spool or log directories while checking their
7359 size for availability, Exim confusingly gave the error "space shortage".
7360 Furthermore, in debugging mode it crashed with a floating point exception.
7361 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7362 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7363 serious problem, Exim now writes to the main and panic logs when this
7364 happens, with details of the failure. It then refuses to accept the
7365 incoming message, giving the message "spool directory problem" or "log
7366 directory problem" with a 421 code for SMTP messages.
7368 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7369 0, 1, and 2 exist, because some OS complain for execs without them (see
7370 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7371 descriptors. However, the code omitted to check that the open succeeded,
7372 causing mysterious errors if for some reason the permissions on /dev/null
7373 got screwed. Now Exim writes a message to the main and panic logs, and
7374 bombs out if it can't open /dev/null.
7376 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7377 interact so that it is all more flexible. It is supposed to remain
7378 backwards compatible. Also added extra_local_interfaces.
7380 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7381 to bomb out with an assertion failure - to the client this appears as a
7382 connection drop. This problem occurs in the part of the code that was taken
7383 from the Samba project. Fortunately, the assertion is in a very simple
7384 function, so I have fixed this by reproducing the function inline in the
7385 one place where it is called, and arranging for authentication to fail
7386 instead of killing the process with assert().
7388 75. The SPA client code was not working when the server requested OEM rather
7389 than Unicode encoding.
7391 76. Added code to make require_files with a specific uid setting more usable in
7392 the case where statting the file as root fails - usually a non-root-mounted
7393 NFS file system. When this happens and the failure is EACCES, Exim now
7394 forks a subprocess and does the per-uid checking as the relevant uid.
7396 77. Added process_log_path.
7398 78. If log_file_path was not explicitly set, a setting of check_log_space or
7399 check_log_inodes was ignored.
7401 79. If a space check for the spool or log partitions fails, the incident is now
7402 logged. Of course, in the latter case the data may get lost...
7404 80. Added the %p formatting code to string_format() so that it can be used to
7405 print addresses in debug_print(). Adjusted all the address printing in the
7406 debugging in store.c to use %p rather than %d.
7408 81. There was a concern that a line of code in smtp_in.c could overflow a
7409 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7410 initially expressed, the concern was not well-founded, because trailing
7411 spaces are removed early. However, if the trailing spaces were followed by
7412 a NULL, they did not get removed, so the overflow was possible. Two fixes
7415 (a) I re-wrote the offending code in a cleaner fashion.
7416 (b) If an incoming SMTP command contains a NULL character, it is rejected
7419 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7420 runs initgroups(), so that if the Exim user is in any additional groups,
7421 they will be used during message reception.
7427 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob