1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
27 JH/01 Remove code calling the customisable local_scan function, unless a new
28 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
30 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
31 non-signal-safe functions being used.
33 JH/03 Bug 2269: When presented with a received message having a stupidly large
34 number of DKIM-Signature headers, disable DKIM verification to avoid
35 a resource-consumption attack. The limit is set at twenty.
37 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
38 report of oldest_pass in ${authres } in consequence, and separate out
39 some descriptions of reasons for verification fail.
41 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
42 files in the spool were present and unlocked. A queue-runner could spot
43 them, resulting in a duplicate delivery. Fix that by doing the unlock
44 after the unlink. Investigation by Tim Stewart. Take the opportunity to
45 add more error-checking on spoolfile handling while that code is being
48 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
49 No known attacks, no CVE, this is defensive hardening.
51 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
52 a queue-runner could start a delivery while other operations were ongoing.
53 Cutthrough delivery was a common victim, resulting in duplicate delivery.
54 Found and investigated by Tim Stewart. Fix by using the open message data
55 file handle rather than opening another, and not locally closing it (which
56 releases a lock) for that case, while creating the temporary .eml format
57 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
59 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
60 $sender_verify_failure/$recipient_verify_failure to "random".
62 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
65 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
66 Previously this would segfault.
68 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
71 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
72 like zero, since the resolver should be doing this for us, But we need one
73 as a CNAME but no MX presence gets the CNAME returned; we need to check
74 that doesn't point to an MX to declare it "no MX returned" rather than
75 "error, loop". A new main option is added so the older capability of
76 following some limited number of chain links is maintained.
78 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
80 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
81 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
82 now-deprecated earlier definitions used only specified the range up to TLS
83 1.2 (in the older-version library docs).
85 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
87 JH/15 Rework TLS client-side context management. Stop using a global, and
88 explicitly pass a context around. This enables future use of TLS for
89 connections to service-daemons (eg. malware scanning) while a client smtp
90 connection is using TLS; with cutthrough connections this is quite likely.
92 JH/16 Fix ARC verification to do AS checks in reverse order.
94 JH/17 Support a "tls" option on the ${readsocket } expansion item.
96 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
97 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
98 Previously the "utf8" would be re-prepended for every additional message.
100 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
101 Previously thery were accepted, resulting in issues when attempting to
102 forward messages to a non-supporting MTA.
104 PP/02 Let -n work with printing macros too, not just options.
106 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
107 one parent address was copied, and bogus data was used at delivery-logging
108 time. Either a crash (after delivery) or bogus log data could result.
109 Discovery and analysis by Tim Stewart.
111 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
112 Previously if the string ended mid-character, we did not insert the
113 promised '?' replacement.
115 PP/04 Documentation: current string operators work on bytes, not codepoints.
117 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
118 should pack well giving a smaller memory footprint so better caching and
119 therefore performance. Group the declarations where this can't be done so
120 that the byte-sized flag variables are not interspersed among pointer
121 variables, giving a better chance of good packing by the compiler.
123 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
124 non-null, to avoid issues with sites running BATV. Previously reports were
125 sent with an empty envelope sender so looked like bounces.
127 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
128 The ignore_error flag wasn't being returned from the filter subprocess so
129 was not set for later routers. Investigation and fix by Matthias Kurz.
131 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
132 and a msg:complete for the whole, when a message is manually removed using
133 -Mrm. Developement by Matthias Kurz, hacked on by JH.
135 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
136 a "Gnu special" function, asprintf() in the DB utility binary builds; I
137 hope that is portable enough.
139 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
140 requiring a known-CA anchor certificate; make it now rely entirely on the
141 TLSA as an anchor. Checking the name on the leaf cert against the name
142 on the A-record for the host is still done for TA (but not for EE mode).
144 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
145 included in delivery lines for non-proxied connections, when compiled with
146 SUPPORT_SOCKS and running with proxy logging enabled.
148 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
149 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
150 move the existing event to fire before the normal logging of message
151 failure so that custom logging is bracketed by normal logging.
153 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
154 msg:fail:internal event. Developement by Matthias Kurz.
156 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
157 far too small for todays use of crypto signatures stored there. Go all
158 the way to the max DNS message size of 64kB, even though this might be
159 overmuch for IOT constrained device use.
161 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
162 copy a string over itself. The library routine is documented as not
163 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
165 JH/32 For main options check_spool_space and check_inode_space, where the
166 platform supports 64b integers, support more than the previous 2^31 kB
167 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
168 the previous G, M, k.
170 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
171 $authenticated_fail_id variable on authentication failure. Previously
174 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
175 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
176 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
177 GNUTLS_SEC_PARAM_MEDIUM.
179 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
180 side. Previously we would continue as if no SNI had been received.
182 JH/36 Harden the handling of string-lists. When a list consisted of a sole
183 "<" character, which should be a list-separator specification, we walked
184 off past the nul-terimation.
186 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
187 causes) even when the retry time is not yet met. Previously they were
188 not, meaning that when (say) an account was over-quota and temp-rejecting,
189 and multiple senders' messages were queued, only one sender would get
190 notified on each configured delay_warning cycle.
192 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
194 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
195 cipher-suites, an error can be left on the stack even for a succeeding
196 accept; this results in impossible error messages when a later operation
199 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
200 return error codes indicating retry. Under TLS1.3 this becomes required.
202 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
203 it only wrote the new authenticators, resulting in a lack of tracking of
204 peer changes of ESMTP extensions until the next cache flush.
206 JH/41 Fix the loop reading a message header line to check for integer overflow,
207 and more-often against header_maxsize. Previously a crafted message could
208 induce a crash of the recive process; now the message is cleanly rejected.
210 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
211 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
217 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
218 When redis_servers is set to a list of > 1 element, and the Redis servers
219 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
220 case of MOVED into a DEFER case instead, thus moving the query onto the
221 next server in the list. For a cluster of N elements, all N servers must
222 be defined in redis_servers.
224 GF/02 Catch and remove uninitialized value warning in exiqsumm
225 Check for existence of @ARGV before looking at $ARGV[0]
227 JH/01 Replace the store_release() internal interface with store_newblock(),
228 which internalises the check required to safely use the old one, plus
229 the allocate and data copy operations duplicated in both (!) of the
230 extant use locations.
232 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
233 modifier. This matches the restriction on the commandline.
235 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
236 Previously only the last row was returned.
238 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
239 we assumed that tags in the header were well-formed, and parsed the
240 element content after inspecting only the first char of the tag.
241 Assumptions at that stage could crash the receive process on malformed
244 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
245 While running the DKIM ACL we operate on the Permanent memory pool so that
246 variables created with "set" persist to the DATA ACL. Also (at any time)
247 DNS lookups that fail create cache records using the Permanent pool. But
248 expansions release any allocations made on the current pool - so a dnsdb
249 lookup expansion done in the DKIM ACL releases the memory used for the
250 DNS negative-cache, and bad things result. Solution is to switch to the
251 Main pool for expansions.
252 While we're in that code, add checks on the DNS cache during store_reset,
253 active in the testsuite.
254 Problem spotted, and debugging aided, by Wolfgang Breyha.
256 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
257 When none of the hosts presented to a transport match an already-open
258 connection, close it and proceed with the list. Previously we would
259 queue the message. Spotted by Lena with Yahoo, probably involving
262 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
263 Previously a spurious "250 OK id=" response was appended to the proper
266 JH/08 The "support for" informational output now, which built with Content
267 Scanning support, has a line for the malware scanner interfaces compiled
268 in. Interface can be individually included or not at build time.
270 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
271 by the template makefile "src/EDITME". The "STREAM" support for an older
272 ClamAV interface method is removed.
274 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
275 rows affected is given instead).
277 JH/11 The runtime Berkeley DB library version is now additionally output by
278 "exim -d -bV". Previously only the compile-time version was shown.
280 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
281 SMTP connection. Previously, when one had more receipients than the
282 first, an abortive onward connection was made. Move to full support for
283 multiple onward connections in sequence, handling cutthrough connection
284 for all multi-message initiating connections.
286 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
287 routers. Previously, a multi-recipient message would fail to match the
288 onward-connection opened for the first recipient, and cause its closure.
290 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
291 a timeout on read on a GnuTLS initiating connection, resulting in the
292 initiating connection being dropped. This mattered most when the callout
293 was marked defer_ok. Fix to keep the two timeout-detection methods
296 JH/15 Relax results from ACL control request to enable cutthrough, in
297 unsupported situations, from error to silently (except under debug)
298 ignoring. This covers use with PRDR, frozen messages, queue-only and
301 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
303 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
304 metadata, resulting in a crash in free().
306 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
307 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
308 Broken also in d185889f4, with init system revamp.
310 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
311 Previously we abruptly closed the connection after reading a malware-
312 found indication; now we go on to read the "scan ok" response line,
315 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
316 ACL. Previously, a crash would result.
318 JH/19 Speed up macro lookups during configuration file read, by skipping non-
319 macro text after a replacement (previously it was only once per line) and
320 by skipping builtin macros when searching for an uppercase lead character.
322 JH/20 DANE support moved from Experimental to mainline. The Makefile control
323 for the build is renamed.
325 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
326 was allocated for every new TLS startup, meaning one per message. Fix
327 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
329 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
330 reported the original. Fix to report (as far as possible) the ACL
331 result replacing the original.
333 JH/23 Fix memory leak during multi-message connections using STARTTLS under
334 OpenSSL. Certificate information is loaded for every new TLS startup,
335 and the resources needed to be freed.
337 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
339 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
342 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
343 DATA response info to the (existing) per-recipient response info for
344 the "C=" log element. It can have useful tracking info from the
345 destination system. Patch from Simon Arlott.
347 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
348 length value. Previously this would segfault.
350 HS/02 Support Avast multiline protoocol, this allows passing flags to
351 newer versions of the scanner.
353 JH/28 Ensure that variables possibly set during message acceptance are marked
354 dead before release of memory in the daemon loop. This stops complaints
355 about them when the debug_store option is enabled. Discovered specifically
356 for sender_rate_period, but applies to a whole set of variables.
357 Do the same for the queue-runner and queue-list loops, for variables set
358 from spool message files. Do the same for the SMTP per-message loop, for
359 certain variables indirectly set in ACL operations.
361 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
362 as a multi-recipient message from a mailinglist manager). The coding had
363 an arbitrary cutoff number of characters while checking for more input;
364 enforced by writing a NUL into the buffer. This corrupted long / fast
365 input. The problem was exposed more widely when more pipelineing of SMTP
366 responses was introduced, and one Exim system was feeding another.
367 The symptom is log complaints of SMTP syntax error (NUL chars) on the
368 receiving system, and refused recipients seen by the sending system
369 (propating to people being dropped from mailing lists).
370 Discovered and pinpointed by David Carter.
372 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
373 replaced by the ${authresults } expansion.
375 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
377 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
378 allows proper process termination in container environments.
380 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
381 Previously the "final dot" had a newline after it; ensure it is CR,LF.
383 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
384 and "err_perm", deprecated since 4.83 when the RFC-defined words
385 "temperror" and "permerror" were introduced.
387 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
388 transport-filters or DKIM-signing. The restriction was lost in the
389 consolidation of verify-callout and delivery SMTP handling.
390 Extend the restriction to also cover ARC-signing.
392 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
393 in defer=pass mode supply a 450 to the initiator. Previously the message
396 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
397 tls_require_ciphers is used as before.
399 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
400 "pass_unscanned". Only tmpfails from the scanner are written to
401 the paniclog, as they may require admin intervention (permission
402 denied, license issues). Other scanner errors (like decompression
403 bombs) do not cause a paniclog entry.
405 JH/36 Fix reinitialisation of DKIM logging variable between messages.
406 Previously it was possible to log spurious information in receive log
409 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
410 triggered odd behaviour from Outlook Express clients.
412 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
415 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
416 since the IETF WG has not yet settled on that versus the original
417 "bare" representation.
419 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
420 Previously the millisecond value corrupted the output.
421 Fix also for syslog_pid=no and log_selector +pid, for which the pid
422 corrupted the output.
428 JH/01 Rework error string handling in TLS interface so that the caller in
429 more cases is responsible for logging. This permits library-sourced
430 string to be attached to addresses during delivery, and collapses
431 pairs of long lines into single ones.
433 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
434 during configuration. Wildcards are allowed and expanded.
436 JH/02 Rework error string handling in DKIM to pass more info back to callers.
437 This permits better logging.
439 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
440 do not close it down and have the child transport start it up again on
441 the passed-on TCP connection. Instead, proxy the child (and any
442 subsequent ones) for TLS via a unix-domain socket channel. Logging is
443 affected: the continued delivery log lines do not have any DNSSEC, TLS
444 Certificate or OCSP information. TLS cipher information is still logged.
446 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
447 identical IP addresses on different listening ports. Will also affect
450 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
451 add noisy ifdef guards to special-case this sillyness.
452 Patch from Bernd Kuhls.
454 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
455 than 255 are no longer allowed.
457 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
458 Disable the session-cache too, which might reduce our load. Since we
459 currrectly use a new context for every connection, both as server and
460 client, there is no benefit for these.
461 GnuTLS appears to not support tickets server-side by default (we don't
462 call gnutls_session_ticket_enable_server()) but client side is enabled
463 by default on recent versions (3.1.3 +) unless the PFS priority string
466 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
467 <https://reproducible-builds.org/specs/source-date-epoch/>.
469 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
470 the check for any unsuccessful recipients did not notice the limit, and
471 erroneously found still-pending ones.
473 JH/08 Pipeline CHUNKING command and data together, on kernels that support
474 MSG_MORE. Only in-clear (not on TLS connections).
476 JH/09 Avoid using a temporary file during transport using dkim. Unless a
477 transport-filter is involved we can buffer the headers in memory for
478 creating the signature, and read the spool data file once for the
479 signature and again for transmission.
481 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
482 4.77 as the kernel support then wasn't solid, having issues in 64bit
483 mode. Now, it's been long enough. Add support for FreeBSD also.
485 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
486 case where the routing stage had gathered several addresses to send to
487 a host before calling the transport for the first, we previously failed
488 to close down TLS in the old transport process before passing the TCP
489 connection to the new process. The new one sent a STARTTLS command
490 which naturally failed, giving a failed delivery and bloating the retry
491 database. Investigation and fix prototype from Wolfgang Breyha.
493 JH/12 Fix check on SMTP command input synchronisation. Previously there were
494 false-negatives in the check that the sender had not preempted a response
495 or prompt from Exim (running as a server), due to that code's lack of
496 awareness of the SMTP input buffering.
498 PP/04 Add commandline_checks_require_admin option.
499 Exim drops privileges sanely, various checks such as -be aren't a
500 security problem, as long as you trust local users with access to their
501 own account. When invoked by services which pass untrusted data to
502 Exim, this might be an issue. Set this option in main configuration
503 AND make fixes to the calling application, such as using `--` to stop
506 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
507 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
508 responses to those, into a single TLS record each way (this usually means
509 a single packet). As a side issue, smtp_enforce_sync now works on TLS
512 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
513 affects you only if you're dancing at the edge of the param size limits.
514 If you are, and this message makes sense to you, then: raise the
515 configured limit or use OpenSSL 1.1. Nothing we can do for older
518 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
519 cmdline element to get the documented default one. Previously it was
522 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
523 are unsuable for non-message-specific reasons - eg. connection timeout,
524 banner-time rejection.
526 JH/16 Fix logging of delivery remote port, when specified by router, under
529 PP/06 Repair manualroute's ability to take options in any order, even if one
530 is the name of a transport.
533 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
535 JH/17 Change the list-building routines interface to use the expanding-string
536 triplet model, for better allocation and copying behaviour.
538 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
539 Previously it was constructed the first time a possibly-matching string
540 was met in the configuration file input during startup; now it is done
543 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
544 compatible one, to avoid the (poorly documented) possibility of a config
545 file in the working directory redirecting the DB files, possibly correpting
546 some existing file. CVE-2017-10140 assigned for BDB.
548 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
549 cache-hot. Previously, although the result was properly cached, the
550 initial verify call returned a defer.
552 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
553 the main verify for receipient in uncached-mode.
555 JH/22 Retire historical build files to an "unsupported" subdir. These are
556 defined as "ones for which we have no current evidence of testing".
558 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
559 if present. Previously it was ignored.
561 JH/24 Start using specified-initialisers in C structure init coding. This is
562 a C99 feature (it's 2017, so now considered safe).
564 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
565 if was a fixed-sized field and bitmask ops via macros; it is now more
568 PP/07 GitHub PR 56: Apply MariaDB build fix.
569 Patch provided by Jaroslav Škarvada.
571 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
572 during Coverity cleanups [4.87 JH/47]
573 Diagnosis and fix provided by Michael Fischer v. Mollard.
575 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
576 the right size to place the terminating semicolon on its own folded
577 line, the header hash was calculated to an incorrect value thanks to
578 the (relaxed) space the fold became.
580 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
581 and confused the parent.
583 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
584 which could crash as a result. This could lead to undeliverable messages.
586 JH/28 Logging: "next input sent too soon" now shows where input was truncated
589 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
590 matters on fast-turnover and PID-randomising systems, which were getting
591 out-of-order delivery.
593 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
594 a possibly-overlapping copy. The symptom was that "Remote host closed
595 connection in response to HELO" was logged instead of the actual 4xx
598 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
599 Previously only that bufferd was discarded, resulting in SYMTP command
602 JH/32 DKIM: when a message has multiple signatures matching an identity given
603 in dkim_verify_signers, run the dkim acl once for each. Previously only
604 one run was done. Bug 2189.
606 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
607 "panic the current process" to "deliberately defer". The panic log is
608 still written with the problem list name; the mail and reject logs now
609 get a temp-reject line for the message that was being handled, saying
610 something like "domains check lookup or other defer". The SMTP 451
611 message is still "Temporary local problem".
613 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
614 A crafted sequence of BDAT commands could result in in-use memory beeing
615 freed. CVE-2017-16943.
617 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
618 from SMTP input. Previously it was always done; now only done for DATA
619 and not BDAT commands. CVE-2017-16944.
621 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
622 to the message (such as an overlong header line). Previously this was
623 not done and we did not exit BDAT mode. Followon from the previous item
624 though a different problem.
630 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
631 than -2003 did; needs libidn2 in addition to libidn.
633 JH/02 The path option on a pipe transport is now expanded before use.
635 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
636 Patch provided by "Björn", documentation fix added too.
638 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
639 missing a wire-to-host endian conversion.
641 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
642 close after a BDAT command line could be taken as a following command,
643 giving a synch failure. Fix by only checking for synch immediately
644 before acknowledging the chunk.
646 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
647 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
648 macro. Patches provided by Josh Soref.
650 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
651 Previously we did not; the RFC seems ambiguous and VRFY is not listed
652 by IANA as a service extension. However, John Klensin suggests that we
655 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
656 the dkim code may be unix-mode line endings rather than smtp wire-format
657 CRLF, so prepend a CR to any bare LF.
659 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
660 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
662 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
663 the first were themselves being wrongly included in the feed into dkim
664 processing; with most chunk sizes in use this resulted in an incorrect
665 body hash calculated value.
667 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
668 DKIM signature block, for verification. Although advised against by
669 standards it is specifically not ruled illegal.
671 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
673 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
674 missing a body hash (the bh= tag).
676 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
677 It seems that HAProxy sends the Proxy Protocol information in clear and
678 only then does a TLS startup, so do the same.
680 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
681 TCP connections (such as for Spamd) unless the daemon successfully set
682 Fast Open mode on its listening sockets. This fixes breakage seen on
683 too-old kernels or those not configured for Fast Open, at the cost of
684 requiring both directions being enabled for TFO, and TFO never being used
685 by non-daemon-related Exim processes.
687 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
688 endings, at least on the first header line. Try to canonify any that get
689 past that check, despite the cost.
691 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
692 now limited to an arbitrary five deep, while parsing addresses with the
693 strip_excess_angle_brackets option enabled.
695 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
696 instead leave the unprompted TLS handshake in socket buffer for the
697 TLS library to consume.
699 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
701 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
703 JH/16 Drop variables when they go out of scope. Memory management drops a whole
704 region in one operation, for speed, and this leaves assigned pointers
705 dangling. Add checks run only under the testsuite which checks all
706 variables at a store-reset and panics on a dangling pointer; add code
707 explicitly nulling out all the variables discovered. Fixes one known
708 bug: a transport crash, where a dangling pointer for $sending_ip_address
709 originally assigned in a verify callout, is re-used.
711 PP/06 Drop '.' from @INC in various Perl scripts.
713 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
715 PP/08 Reduce a number of compilation warnings under clang; building with
716 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
717 should be warning-free.
719 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
721 HS/01 Fix portability problems introduced by PP/08 for platforms where
722 realloc(NULL) is not equivalent to malloc() [SunOS et al].
724 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
725 chunk. This allows us to accept broken chunked messages. We need a more
726 general solution here.
728 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
729 already-broken messages in the queue.
731 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
733 JH/19 Fix reference counting bug in routing-generated-address tracking.
739 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
740 supports it and a size is available (ie. the sending peer gave us one).
742 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
743 years of being deprecated). The replacements are the ACLs
744 acl_smtp_mime and acl_not_smtp_mime.
746 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
747 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
748 or even an in-clear connection were permitted. Now, if the host lookup
749 was dnssec and dane was requested then the host is only used if the
750 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
751 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
752 if one fails this test.
753 This means that a poorly-configured remote DNS will make it incommunicado;
754 but it protects against a DNS-interception attack on it.
756 JH/04 Bug 1810: make continued-use of an open smtp transport connection
757 non-noisy when a race steals the message being considered.
759 JH/05 If main configuration option tls_certificate is unset, generate a
760 self-signed certificate for inbound TLS connections.
762 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
763 in rewrites and routers.
765 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
766 and logged a warning sing 4.83; now they are a configuration file error.
768 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
769 (lacking @domain). Apply the same qualification processing as RCPT.
771 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
773 JH/10 Support ${sha256:} applied to a string (as well as the previous
776 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
777 a cutthrough deliver is pending, as we always want to make a connection.
778 This also avoids re-routing the message when later placing the cutthrough
779 connection after a verify cache hit.
780 Do not update it with the verify result either.
782 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
783 when routing results in more than one destination address.
785 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
786 signing (which inhibits the cutthrough capability). Previously only
787 the presence of an option was tested; now an expansion evaluating as
788 empty is permissible (obviously it should depend only on data available
789 when the cutthrough connection is made).
791 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
792 the relevant preceding SMTP command did not note the pipelining mode.
794 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
795 Previously they were not counted.
797 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
798 as one having no matching records. Previously we deferred the message
799 that needed the lookup.
801 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
802 distinguished as "(=".
804 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
805 for missing MX records. Previously it only worked for missing A records.
807 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
809 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
810 after the data-go-ahead and data-ack. Patch from Jason Betts.
812 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
813 even for a "none" policy. Patch from Tony Meyer.
815 JH/22 Fix continued use of a connection for further deliveries. If a port was
816 specified by a router, it must also match for the delivery to be
819 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
820 When one of the recipients of a message was unsuitable for the connection
821 (has no matching addresses), we lost track of needing to mark it
822 deferred. As a result mail would be lost.
824 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
826 JH/25 Decoding ACL controls is now done using a binary search; the source code
827 takes up less space and should be simpler to maintain. Merge the ACL
828 condition decode tables also, with similar effect.
830 JH/26 Fix problem with one_time used on a redirect router which returned the
831 parent address unchanged. A retry would see the parent address marked as
832 delivered, so not attempt the (identical) child. As a result mail would
835 JH/27 Fix a possible security hole, wherein a process operating with the Exim
836 UID can gain a root shell. Credit to http://www.halfdog.net/ for
837 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
840 JH/28 Enable {spool,log} filesystem space and inode checks as default.
841 Main config options check_{log,spool}_{inodes,space} are now
842 100 inodes, 10MB unless set otherwise in the configuration.
844 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
845 Previously it only applied to the main-section connection policy
848 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
850 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
851 by me. Added RFC7919 DH primes as an alternative.
853 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
854 are not in the system include path.
856 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
857 GnuTLS, when a session startup failed (eg because the client disconnected)
858 Exim did stdio operations after fclose. This was exposed by a recent
859 change which nulled out the file handle after the fclose.
861 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
862 signed directly by the cert-signing cert, rather than an intermediate
863 OCSP-signing cert. This is the model used by LetsEncrypt.
865 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
867 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
868 an incoming connection.
870 HS/02 Bug 1802: Do not half-close the connection after sending a request
873 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
874 fallback to "prime256v1".
876 JH/34 SECURITY: Use proper copy of DATA command in error message.
877 Could leak key material. Remotely exploitable. CVE-2016-9963.
883 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
884 and 3.4.4 - once the server is enabled to respond to an OCSP request
885 it does even when not requested, resulting in a stapling non-aware
886 client dropping the TLS connection.
888 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
889 support variable-length bit vectors. No functional change.
891 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
892 The I= interface field on outgoing lines is now after the H= remote
893 host field, same as incoming lines. There is a separate
894 outgoing_interface log selector which allows you to disable the
897 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
898 If not running log_selector +smtp_connection the mainlog would be held
899 open indefinitely after a "too many connections" event, including to a
900 deleted file after a log rotate. Leave the per net connection logging
901 leaving it open for efficiency as that will be quickly detected by the
902 check on the next write.
904 HS/01 Bug 1671: Fix post transport crash.
905 Processing the wait-<transport> messages could crash the delivery
906 process if the message IDs didn't exist for some reason. When
907 using 'split_spool_directory=yes' the construction of the spool
908 file name failed already, exposing the same netto behaviour.
910 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
911 mime_regex ACL conditions.
913 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
914 to DSN fail messages (bounces): remote IP, remote greeting, remote response
915 to HELO, local diagnostic string.
917 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
918 log line to debug. Even when configured with a tls authenticator many
919 client connections are expected to not authenticate in this way, so
920 an authenticate fail is not an error.
922 HS/02 Add the Exim version string to the process info. This way exiwhat
923 gives some more detail about the running daemon.
925 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
926 matter for fast-change records such as DNSBLs.
928 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
929 retry record, even if constant. There may be multiple transports with
930 different interface settings and the retry behaviour needs to be kept
933 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
935 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
937 JH/10 Bug 840: fix log_defer_output option of pipe transport
939 JH/11 Bug 830: use same host for all RCPTS of a message, even under
940 hosts_randomize. This matters a lot when combined with mua_wrapper.
942 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
943 ${quote_pgsql:<string>} operator.
945 JH/13 Bug 1708: avoid misaligned access in cached lookup.
947 JH/14 Change header file name for freeradius-client. Relevant if compiling
948 with Radius support; from the Gentoo tree and checked under Fedora.
950 JH/15 Bug 1712: Introduce $prdr_requested flag variable
952 JH/16 Bug 1714: Permit an empty string as expansion result for transport
953 option transport_filter, meaning no filtering.
955 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
957 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
958 defaults to "*" (all hosts). The variable is now available when not built
959 with TLS, default unset, mainly to enable keeping the testsuite sane.
960 If a server certificate is not supplied (via tls_certificate) an error is
961 logged, and clients will find TLS connections fail on startup. Presumably
962 they will retry in-clear.
963 Packagers of Exim are strongly encouraged to create a server certificate
964 at installation time.
966 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
967 with the $config_file variable.
969 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
970 in transport context, after the attempt, and per-recipient. The latter type
971 is per host attempted. The event data is the error message, and the errno
972 information encodes the lookup type (A vs. MX) used for the (first) host,
973 and the trailing two digits of the smtp 4xx response.
975 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
976 to write to mainlog (or rejectlog, paniclog) in the window between file
977 creation and permissions/ownership being changed. Particularly affects
978 installations where exicyclog is run as root, rather than exim user;
979 result is that the running daemon panics and dies.
981 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
983 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
984 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
985 "pri" and "weight". Note that the previous implicit priority given by the
986 list order is no longer honoured.
988 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
991 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
992 by defining SUPPORT_SOCKS.
994 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
995 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
996 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
997 variables are renamed to proxy_{local,external}_{address,port}.
999 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1000 for a build by defining SUPPORT_I18N
1002 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1003 of the query string, and make ${quote_redis:} do that quoting.
1005 JH/29 Move Events support from Experimental to mainline, enabled by default
1006 and removable for a build by defining DISABLE_EVENT.
1008 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1010 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1011 cached by the daemon.
1013 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1014 by defining LOOKUP_REDIS. The libhiredis library is required.
1016 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1017 keys are given for lookup.
1019 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1020 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1021 only supported when built with TLS support. The PolarSSL SHA routines
1022 are still used when the TLS library is too old for convenient support.
1024 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1025 openssl_options), for security. OpenSSL forces this from version 1.1.0
1026 server-side so match that on older versions.
1028 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1029 allocation for $value could be released as the expansion processing
1030 concluded, but leaving the global pointer active for it.
1032 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1033 and to use the domains and local_parts ACL conditions.
1035 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1036 incorrectly not doubled on cutthrough transmission, hence seen as a
1037 body-termination at the receiving system - resulting in truncated mails.
1038 Commonly the sender saw a TCP-level error, and retransmitted the message
1039 via the normal store-and-forward channel. This could result in duplicates
1040 received - but deduplicating mailstores were liable to retain only the
1041 initial truncated version.
1043 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1045 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1047 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1048 we're in there, support oversigning also; bug 1309.
1050 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1052 HS/04 Add support for keep_environment and add_environment options.
1054 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1055 either intentional arithmetic overflow during PRNG, or testing config-
1058 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1059 delivery resulted in actual delivery. Cancel cutthrough before DATA
1062 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1063 rejects a recipient: pass the reject to the originator.
1065 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1066 Many were false-positives and ignorable, but it's worth fixing the
1069 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1070 for the new environment-manipulation done at startup. Move the routines
1071 from being local to tls.c to being global via the os.c file.
1073 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1074 an extract embedded as result-arg for a map, the first arg for extract
1075 is unavailable so we cannot tell if this is a numbered or keyed
1076 extraction. Accept either.
1082 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1085 JH/02 The smtp transport option "multi_domain" is now expanded.
1087 JH/03 The smtp transport now requests PRDR by default, if the server offers
1090 JH/04 Certificate name checking on server certificates, when exim is a client,
1091 is now done by default. The transport option tls_verify_cert_hostnames
1092 can be used to disable this per-host. The build option
1093 EXPERIMENTAL_CERTNAMES is withdrawn.
1095 JH/05 The value of the tls_verify_certificates smtp transport and main options
1096 default to the word "system" to access the system default CA bundle.
1097 For GnuTLS, only version 3.0.20 or later.
1099 JH/06 Verification of the server certificate for a TLS connection is now tried
1100 (but not required) by default. The verification status is now logged by
1101 default, for both outbound TLS and client-certificate supplying inbound
1104 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1107 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1108 Status Notification (bounce) messages are now MIME format per RFC 3464.
1109 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1110 under the control of the dsn_advertise_hosts option, and routers may
1111 have a dsn_lasthop option.
1113 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1114 default, modifiable by a malware= option. The list separator for
1115 the options can now be changed in the usual way. Bug 68.
1117 JH/10 The smtp_receive_timeout main option is now expanded before use.
1119 JH/11 The incoming_interface log option now also enables logging of the
1120 local interface on delivery outgoing connections.
1122 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1123 if the interface and destination host and port all match.
1125 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1128 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1129 Patch from Andrew Lewis.
1131 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1132 now supports optional time-restrictions, weighting, and priority
1133 modifiers per server. Patch originally by <rommer@active.by>.
1135 JH/16 The spamd_address main option now supports a mixed list of local
1136 and remote servers. Remote servers can be IPv6 addresses, and
1137 specify a port-range.
1139 JH/17 Bug 68: The spamd_address main option now supports an optional
1140 timeout value per server.
1142 JH/18 Bug 1581: Router and transport options headers_add/remove can
1143 now have the list separator specified.
1145 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1148 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1151 JH/21 Support for the A6 type of dns record is withdrawn.
1153 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1154 rather than the verbs used.
1156 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1157 from 255 to 1024 chars.
1159 JH/24 Verification callouts now attempt to use TLS by default.
1161 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1162 are generic router options now. The defaults didn't change.
1164 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1165 Original patch from Alexander Shikoff, worked over by JH.
1167 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1168 files if ZCAT_COMMAND is not executable.
1170 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1172 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1174 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1175 Normally benign, it bites when the pair was led to by a CNAME;
1176 modern usage is to not canonicalize the domain to a CNAME target
1177 (and we were inconsistent anyway for A-only vs AAAA+A).
1179 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1181 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1182 when evaluating $sender_host_dnssec.
1184 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1185 $sender_helo_dnssec variable.
1187 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1189 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1191 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1193 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1194 documented as working, but never had. Support all but $spam_report.
1196 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1197 added for tls authenticator.
1199 HS/03 Add perl_taintmode main config option
1205 TL/01 When running the test suite, the README says that variables such as
1206 no_msglog_check are global and can be placed anywhere in a specific
1207 test's script, however it was observed that placement needed to be near
1208 the beginning for it to behave that way. Changed the runtest perl
1209 script to read through the entire script once to detect and set these
1210 variables, reset to the beginning of the script, and then run through
1211 the script parsing/test process like normal.
1213 TL/02 The BSD's have an arc4random API. One of the functions to induce
1214 adding randomness was arc4random_stir(), but it has been removed in
1215 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1216 function when detected.
1218 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1219 cause callback expansion.
1221 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1222 syntax errors in an expansion can be treated as a string instead of
1223 logging or causing an error, due to the internal use of bool_lax
1224 instead of bool when processing it.
1226 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1227 server certificates when making smtp deliveries.
1229 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1231 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1233 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1235 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1236 Merged patch from Sebastian Wiedenroth.
1238 JH/05 Fix results-pipe from transport process. Several recipients, combined
1239 with certificate use, exposed issues where response data items split
1240 over buffer boundaries were not parsed properly. This eventually
1241 resulted in duplicates being sent. This issue only became common enough
1242 to notice due to the introduction of connection certificate information,
1243 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1245 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1246 size buffer was used, resulting in syntax errors when an expansion
1249 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1250 version 3.3.6 or later.
1252 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1253 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1254 both become "event_action", the variables become $event_name, $event_data
1255 and $event_defer_errno. There is a new variable $verify_mode, usable in
1256 routers, transports and related events. The tls:cert event is now also
1257 raised for inbound connections, if the main configuration event_action
1260 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1261 early OCSP support, but no stapling (appears to be less than 1.0.0).
1263 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1264 server certificate names available under the smtp transport option
1265 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1268 JH/10 Time-related extraction expansions from certificates now use the main
1269 option "timezone" setting for output formatting, and are consistent
1270 between OpenSSL and GnuTLS compilations. Bug 1541.
1272 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1273 encoded parameter in the incoming message. Bug 1558.
1275 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1276 include certificate info, eximon was claiming there were spoolfile
1279 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1281 JH/14 Log delivery-related information more consistently, using the sequence
1282 "H=<name> [<ip>]" wherever possible.
1284 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1285 are problematic for Debian distribution, omit them from the release
1288 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1290 JH/16 Fix string representation of time values on 64bit time_t architectures.
1293 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1294 output list separator was used.
1299 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1300 checkers that were complaining about end of non-void function with no
1303 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1304 This was a regression introduced in 4.83 by another bugfix.
1306 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1308 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1309 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1315 TF/01 Correctly close the server side of TLS when forking for delivery.
1317 When a message was received over SMTP with TLS, Exim failed to clear up
1318 the incoming connection properly after forking off the child process to
1319 deliver the message. In some situations the subsequent outgoing
1320 delivery connection happened to have the same fd number as the incoming
1321 connection previously had. Exim would try to use TLS and fail, logging
1322 a "Bad file descriptor" error.
1324 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1325 utilities have not been installed.
1327 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1328 temporary space as the ACL may create new global variables.
1330 TL/01 LDAP support uses per connection or global context settings, depending
1331 upon the detected version of the libraries at build time.
1333 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1334 to extract and use the src ip:port in logging and expansions as if it
1335 were a direct connection from the outside internet. PPv2 support was
1336 updated based on HAProxy spec change in May 2014.
1338 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1340 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1341 Properly escape header and check for NULL return.
1343 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1346 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1348 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1349 characters in header names, implemented as a verify condition.
1350 Contributed by Michael Fischer v. Mollard.
1352 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1353 results permerror and temperror. Previous values are deprecated but
1354 still accepted. In a future release, err_perm and err_temp will be
1355 completely removed, which will be a backward incompatibility if the
1356 ACL tests for either of these two old results. Patch contributed by
1357 user bes-internal on the mailing list.
1359 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1361 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1362 selectors, in both main and reject logs.
1364 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1367 JH/07 Add malware type "sock" for talking to simple daemon.
1369 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1371 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1372 routers/transports under cutthrough routing.
1374 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1375 numbers. Touch up "bool" conditional to keep the same definition.
1377 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1379 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1381 JH/12 Expand items in router/transport headers_add or headers_remove lists
1382 individually rather than the list as a whole. Bug 1452.
1384 Required for reasonable handling of multiple headers_ options when
1385 they may be empty; requires that headers_remove items with embedded
1386 colons must have them doubled (or the list-separator changed).
1388 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1389 view the policy declared in the DMARC record. Currently, $dmarc_status
1390 is a combined value of both the record presence and the result of the
1393 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1395 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1396 dnslookup router and the smtp transport (applying to the forward
1399 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1400 of ldap servers used for a specific lookup. Patch provided by Heiko
1403 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1404 New variable $lookup_dnssec_authenticated for observability.
1406 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1407 Patch submitted by Lars Timman.
1409 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1411 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1412 Requires trusted mode and valid format message id, aborts otherwise.
1413 Patch contributed by Heiko Schlichting.
1415 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1416 certextract with support for various fields. Bug 1358.
1418 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1419 is requested by default, modifiable by smtp transport option
1422 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1423 operate on certificate variables to give certificate fingerprints
1424 Also new ${sha256:cert_variable}.
1426 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1428 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1429 Christian Aistleitner.
1431 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1433 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1434 file. Patch from Wolfgang Breyha.
1436 JH/25 Expand the coverage of the delivery $host and $host_address to
1437 client authenticators run in verify callout. Bug 1476.
1439 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1440 align with daemon_smtp_ports. Bug 72.
1442 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1443 support and error reporting did not work properly.
1445 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1446 and is readable. Patch from Andrew Colin Kissa.
1448 TL/14 Enhance documentation of ${run expansion and how it parses the
1449 commandline after expansion, particularly in the case when an
1450 unquoted variable expansion results in an empty value.
1452 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1454 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1455 by renaming away from it. Addresses GH issue 15, reported by
1458 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1460 TL/15 SECURITY: prevent double expansion in math comparison functions
1461 (can expand unsanitized data). Not remotely exploitable.
1468 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1470 PP/02 Make -n do something, by making it not do something.
1471 When combined with -bP, the name of an option is not output.
1473 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1476 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1477 $sender_host_name and config options to manage this, and basic check
1480 PP/05 DSCP support for outbound connections and control modifier for inbound.
1482 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1483 (Only plugin which currently uses this is kerberos4, which nobody should
1484 be using, but we should make it available and other future plugins might
1485 conceivably use it, even though it would break NAT; stuff *should* be
1486 using channel bindings instead).
1488 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1489 name; added for Sendmail compatibility; requires admin caller.
1490 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1491 just ignore it); requires trusted caller.
1492 Also parse but ignore: -Ac -Am -X<logfile>
1495 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1497 TL/02 Add +smtp_confirmation as a default logging option.
1499 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1500 Patch by Magnus Holmgren from 2007-02-20.
1502 TL/04 Bugzilla 1281 - Spec typo.
1503 Bugzilla 1283 - Spec typo.
1504 Bugzilla 1290 - Spec grammar fixes.
1506 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1508 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1510 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1511 mailing list by by Dmitry Isaikin.
1513 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1515 JH/02 Support "G" suffix to numbers in ${if comparisons.
1517 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1519 NM/01 Bugzilla 1197 - Spec typo
1520 Bugzilla 1196 - Spec examples corrections
1522 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1524 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1525 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1528 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1529 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1531 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1532 "acl {{name}{arg}...}", and optional args on acl condition
1535 JH/05 Permit multiple router/transport headers_add/remove lines.
1537 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1539 JH/07 Avoid using a waiting database for a single-message-only transport.
1540 Performance patch from Paul Fisher. Bugzilla 1262.
1542 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1545 JH/09 Add $headers_added variable, with content from use of ACL modifier
1546 add_header (but not yet added to the message). Bugzilla 199.
1548 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1549 Pulled from Bugzilla 817 by Wolfgang Breyha.
1551 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1553 (nb: this is the same fix as in Exim 4.80.1)
1555 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1558 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1561 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1563 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1564 advertise SMTP AUTH mechanism to us, instead of a generic
1565 protocol violation error. Also, make Exim more robust to bad
1566 data from the Dovecot auth socket.
1568 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1570 When a queue runner is handling a message, Exim first routes the
1571 recipient addresses, during which it prunes them based on the retry
1572 hints database. After that it attempts to deliver the message to
1573 any remaining recipients. It then updates the hints database using
1576 So if a recipient address works intermittently, it can get repeatedly
1577 deferred at routing time. The retry hints record remains fresh so the
1578 address never reaches the final cutoff time.
1580 This is a fairly common occurrence when a user is bumping up against
1581 their storage quota. Exim had some logic in its local delivery code
1582 to deal with this. However it did not apply to per-recipient defers
1583 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1585 This change adds a proper retry rule check during routing so that the
1586 final cutoff time is checked against the message's age. We only do
1587 this check if there is an address retry record and there is not a
1588 domain retry record; this implies that previous attempts to handle
1589 the address had the retry_use_local_parts option turned on. We use
1590 this as an approximation for the destination being like a local
1591 delivery, as in LMTP.
1593 I suspect this new check makes the old local delivery cutoff check
1594 redundant, but I have not verified this so I left the code in place.
1596 TF/02 Correct gecos expansion when From: is a prefix of the username.
1598 Test 0254 submits a message to Exim with the header
1602 When I ran the test suite under the user fanf2, Exim expanded
1603 the header to contain my full name, whereas it should have added
1604 a Resent-Sender: header. It erroneously treats any prefix of the
1605 username as equal to the username.
1607 This change corrects that bug.
1609 GF/01 DCC debug and logging tidyup
1610 Error conditions log to paniclog rather than rejectlog.
1611 Debug lines prefixed by "DCC: " to remove any ambiguity.
1613 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1615 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1616 Bug spotted by Jeremy Harris; was flawed since initial commit.
1617 Would have resulted in OCSP responses post-SNI triggering an Exim
1618 NULL dereference and crash.
1620 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1622 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1623 Bug detection, analysis and fix by Samuel Thibault.
1624 Bugzilla 1331, Debian bug #698092.
1626 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1628 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1629 Server implementation by Todd Lyons, client by JH.
1630 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1631 config variable "prdr_enable" controls whether the server
1632 advertises the facility. If the client requests PRDR a new
1633 acl_data_smtp_prdr ACL is called once for each recipient, after
1634 the body content is received and before the acl_smtp_data ACL.
1635 The client is controlled by both of: a hosts_try_prdr option
1636 on the smtp transport, and the server advertisement.
1637 Default client logging of deliveries and rejections involving
1638 PRDR are flagged with the string "PRDR".
1640 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1641 fclose(). Diagnosis by Todd Lyons.
1643 PP/17 Update configure.default to handle IPv6 localhost better.
1644 Patch by Alain Williams (plus minor tweaks).
1647 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1648 This is now consistent with GnuTLS, and is now documented: the
1649 previous undocumented portable approach to treating the option as
1650 unset was to force an expansion failure. That still works, and
1651 an empty string is now equivalent.
1653 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1654 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1655 not performing validation itself.
1657 PP/20 Added force_command boolean option to pipe transport.
1658 Patch from Nick Koston, of cPanel Inc.
1660 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1663 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1665 PP/21 Fix eximon continuous updating with timestamped log-files.
1666 Broken in a format-string cleanup in 4.80, missed when I repaired the
1667 other false fix of the same issue.
1668 Report and fix from Heiko Schlichting.
1671 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1672 Report from Prashanth Katuri.
1674 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1675 It's SecureTransport, so affects any MacOS clients which use the
1676 system-integrated TLS libraries, including email clients.
1678 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1679 using a MIME ACL for non-SMTP local injection.
1680 Report and assistance in diagnosis by Warren Baker.
1682 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1684 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1686 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1687 last id that failed so it may be referenced in subsequent ACL's.
1689 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1692 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1693 ldap library initialization, allowing self-signed CA's to be
1694 used. Also properly sets require_cert option later in code by
1695 using NULL (global ldap config) instead of ldap handle (per
1696 session). Bug diagnosis and testing by alxgomz.
1698 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1699 the src/util/ subdirectory.
1701 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1702 renamed to Transport Post Delivery Action by Jeremy Harris, as
1705 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1706 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1707 redis_servers = needs to be configured which will be used by the redis
1708 lookup. Patch from Warren Baker, of The Packet Hub.
1710 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1712 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1713 hostname or reverse DNS when processing a host list. Used suggestions
1714 from multiple comments on this bug.
1716 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1718 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1719 Missed a few lines, added it to make the runtest require no keyboard
1722 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1723 contains upper case chars. Make router use caseful_local_part.
1725 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1726 support when GnuTLS has been built with p11-kit.
1732 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1734 This, or similar/improved, will also be change PP/11 of 4.82.
1740 PP/01 Handle short writes when writing local log-files.
1741 In practice, only affects FreeBSD (8 onwards).
1742 Bugzilla 1053, with thanks to Dmitry Isaikin.
1744 NM/01 Bugzilla 949 - Documentation tweak
1746 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1749 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1751 PP/02 Implemented gsasl authenticator.
1753 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1755 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1756 `pkg-config foo` for cflags/libs.
1758 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1759 with rest of GSASL and with heimdal_gssapi.
1761 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1762 `pkg-config foo` for cflags/libs for the TLS implementation.
1764 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1765 properties get this fed in as external SSF. A number of robustness
1766 and debugging improvements to the cyrus_sasl authenticator.
1768 PP/08 cyrus_sasl server now expands the server_realm option.
1770 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1771 Patch by Jeremy Harris.
1773 PP/10 Added dbmjz lookup type.
1775 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1777 PP/12 MAIL args handles TAB as well as SP, for better interop with
1778 non-compliant senders.
1779 Analysis and variant patch by Todd Lyons.
1781 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1782 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1783 Patch from Dirk Mueller <dmueller@suse.com>
1785 PP/13 tls_peerdn now print-escaped for spool files.
1786 Observed some $tls_peerdn in wild which contained \n, which resulted
1787 in spool file corruption.
1789 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1790 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1791 or write after TLS renegotiation, which otherwise led to messages
1794 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1795 as a tracking header (ie: a signed header comes before the signature).
1796 Patch from Wolfgang Breyha.
1798 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1799 comma-sep list; embedded commas doubled.
1801 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1803 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1805 Report and patch from Dmitry Banschikov.
1807 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1808 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1809 failures. We appear to now support TLS1.1+ with Exim.
1811 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1812 lets Exim select keys and certificates based upon TLS SNI from client.
1813 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1814 before an outbound SMTP session. New log_selector, +tls_sni.
1816 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1817 NULL dereference. Report and patch from Alun Jones.
1819 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1820 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1821 Not seeing resolver debug output on NetBSD, but suspect this is a
1822 resolver implementation change.
1824 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1825 Left warnings. Added "eximon gdb" invocation mode.
1827 PP/21 Defaulting "accept_8bitmime" to true, not false.
1829 PP/22 Added -bw for inetd wait mode support.
1831 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1832 locate the relevant includes and libraries. Made this the default.
1834 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1835 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1837 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1838 This may cause build issues on older platforms.
1840 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1841 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1842 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1843 Added SNI support via GnuTLS too.
1844 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1846 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1848 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1849 Applied second patch from Janne, implementing suggestion to default
1850 multiple-strings-in-record handling to match SPF spec.
1852 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1854 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1855 read-only, out of scope).
1856 Patch from Wolfgang Breyha, report from Stuart Northfield.
1858 PP/29 Fix three issues highlighted by clang analyser static analysis.
1859 Only crash-plausible issue would require the Cambridge-specific
1860 iplookup router and a misconfiguration.
1861 Report from Marcin Mirosław.
1863 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1865 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1866 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1867 As part of this, removing so much warning spew let me fix some minor
1868 real issues in debug logging.
1870 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1871 assignment on my part. Fixed.
1873 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1874 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1875 Janne Snabb (who went above and beyond: thank you).
1877 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1878 string otherwise requires a connection and a bunch more work and it's
1879 relatively easy to get wrong. Should also expose TLS library linkage
1882 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1883 64-bit ${eval} (JH/03).
1885 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1886 GNU libc to support some of the 64-bit stuff, should not lead to
1887 conflicts. Defined before os.h is pulled in, so if a given platform
1888 needs to override this, it can.
1890 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1891 protection layer was required, which is not implemented.
1892 Bugzilla 1254, patch from Wolfgang Breyha.
1894 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1895 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1896 tls_dhparam take prime identifiers. Also unbreak combination of
1897 OpenSSL+DH_params+TLSSNI.
1899 PP/39 Disable SSLv2 by default in OpenSSL support.
1905 PP/01 Solaris build fix for Oracle's LDAP libraries.
1906 Bugzilla 1109, patch from Stephen Usher.
1908 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1910 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1913 TF/02 Fix a couple more cases where we did not log the error message
1914 when unlink() failed. See also change 4.74-TF/03.
1916 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1917 lock up or crash if it happened to be inside a call to libc when it
1918 got a SIGUSR1 from exiwhat.
1920 The SIGUSR1 handler appends the current process status to the process
1921 log which is later printed by exiwhat. It used to use the general
1922 purpose logging code to do this, but several functions it calls are
1923 not safe for signals.
1925 The new output code in the SIGUSR1 handler is specific to the process
1926 log, and simple enough that it's easy to inspect for signal safety.
1927 Removing some special cases also simplifies the general logging code.
1928 Removing the spurious timestamps from the process log simplifies
1931 TF/04 Improved ratelimit ACL condition.
1933 The /noupdate option has been deprecated in favour of /readonly which
1934 has clearer semantics. The /leaky, /strict, and /readonly update modes
1935 are mutually exclusive. The update mode is no longer included in the
1936 database key; it just determines when the database is updated. (This
1937 means that when you upgrade Exim will forget old rate measurements.)
1939 Exim now checks that the per_* options are used with an update mode that
1940 makes sense for the current ACL. For example, when Exim is processing a
1941 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1942 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1943 must specify per_mail/readonly. If you omit the update mode it defaults to
1944 /leaky where that makes sense (as before) or /readonly where required.
1946 The /noupdate option is now undocumented but still supported for
1947 backwards compatibility. It is equivalent to /readonly except that in
1948 ACLs where /readonly is required you may specify /leaky/noupdate or
1949 /strict/noupdate which are treated the same as /readonly.
1951 A useful new feature is the /count= option. This is a generalization
1952 of the per_byte option, so that you can measure the throughput of other
1953 aggregate values. For example, the per_byte option is now equivalent
1954 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1956 The per_rcpt option has been generalized using the /count= mechanism
1957 (though it's more complicated than the per_byte equivalence). When it is
1958 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1959 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1960 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1961 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1962 Note that using per_rcpt with a non-readonly update mode in more than
1963 one ACL will cause the recipients to be double-counted. (The per_mail
1964 and per_byte options don't have this problem.)
1966 The handling of very low rates has changed slightly. If the computed rate
1967 is less than the event's count (usually one) then this event is the first
1968 after a long gap. In this case the rate is set to the same as this event's
1969 count, so that the first message of a spam run is counted properly.
1971 The major new feature is a mechanism for counting the rate of unique
1972 events. The new per_addr option counts the number of different
1973 recipients that someone has sent messages to in the last time period. It
1974 behaves like per_rcpt if all the recipient addresses are different, but
1975 duplicate recipient addresses do not increase the measured rate. Like
1976 the /count= option this is a general mechanism, so the per_addr option
1977 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1978 example, measure the rate that a client uses different sender addresses
1979 with the options per_mail/unique=$sender_address. There are further
1980 details in the main documentation.
1982 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1984 TF/06 Removed a few PCRE remnants.
1986 TF/07 Automatically extract Exim's version number from tags in the git
1987 repository when doing development or release builds.
1989 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1990 Bugzilla 879. Patch from Paul Fisher.
1992 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1993 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1996 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1998 PP/05 Variable $av_failed, true if the AV scanner deferred.
1999 Bugzilla 1078. Patch from John Horne.
2001 PP/06 Stop make process more reliably on build failure.
2002 Bugzilla 1087. Patch from Heiko Schlittermann.
2004 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2005 Bugzilla 1089. Patch from Heiko Schlittermann.
2007 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2008 Bugzilla 1131. Patch from Holger Weiß.
2010 PP/09 Handle IPv6 addresses with SPF.
2011 Bugzilla 860. Patch from Wolfgang Breyha.
2013 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2015 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2018 PP/11 match_* no longer expand right-hand-side by default.
2019 New compile-time build option, EXPAND_LISTMATCH_RHS.
2020 New expansion conditions, "inlist", "inlisti".
2022 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2024 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2026 PP/14 fix log_write() format string regression from TF/03.
2027 Bugzilla 1152. Patch from Dmitry Isaikin.
2033 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2035 PP/02 Harmonised TLS library version reporting; only show if debugging.
2036 Layout now matches that introduced for other libraries in 4.74 PP/03.
2038 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2040 PP/04 New "dns_use_edns0" global option.
2042 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2045 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2046 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2048 TK/01 Updated PolarSSL code to 0.14.2.
2049 Bugzilla 1097. Patch from Andreas Metzler.
2051 PP/07 Catch divide-by-zero in ${eval:...}.
2052 Fixes bugzilla 1102.
2054 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2057 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2058 format-string attack -- SECURITY: remote arbitrary code execution.
2060 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2061 time unintentionally subject to list matching rules, letting the header
2062 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2063 arbitrary string expansion). This allowed for information disclosure.
2065 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2066 INT_MIN/-1 -- value coerced to INT_MAX.
2072 NM/01 Workaround for PCRE version dependency in version reporting
2075 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2076 This fixes portability to compilers other than gcc, notably
2077 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2079 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2080 makefiles for portability to HP-UX and POSIX correctness.
2082 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2083 Also via indented variable definition in the Makefile.
2084 (Debugging by Oliver Heesakkers).
2086 PP/02 Restore caching of spamd results with expanded spamd_address.
2087 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2089 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2090 Improves build reliability. Fix from: Frank Elsner
2092 NM/02 Fix wide character breakage in the rfc2047 coding
2093 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2095 NM/03 Allow underscore in dnslist lookups
2096 Fixes bug 1026. Patch from Graeme Fowler
2098 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2099 Code patches from Adam Ciarcinski of NetBSD.
2101 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2104 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2105 is logged, to avoid truncation. Patch from John Horne.
2107 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2108 Patch from Jakob Hirsch.
2110 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2111 SQL string expansion failure details.
2112 Patch from Andrey Oktyabrski.
2114 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2115 Patch from Simon Arlott.
2117 PP/09 New lookups functionality failed to compile on old gcc which rejects
2118 extern declarations in function scope.
2119 Patch from Oliver Fleischmann
2121 PP/10 Use sig_atomic_t for flags set from signal handlers.
2122 Check getgroups() return and improve debugging.
2123 Fixed developed for diagnosis in bug 927 (which turned out to be
2126 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2127 Patch from Mark Zealey.
2129 PP/12 Bugzilla 1056: Improved spamd server selection.
2130 Patch from Mark Zealey.
2132 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2133 Based on patch from Heiko Schlittermann.
2135 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2136 Patch from Uwe Doering, sign-off by Michael Haardt.
2138 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2139 variable declaration deep within a block. Bug and patch from
2142 PP/15 lookups-Makefile IRIX compatibility coercion.
2144 PP/16 Make DISABLE_DKIM build knob functional.
2146 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2147 Patch by Simon Arlott
2149 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2150 variable argument macros. Our copy now differs from upstream.
2156 TF/01 Failure to get a lock on a hints database can have serious
2157 consequences so log it to the panic log.
2159 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2160 controlled using the smtp_confirmation log selector.
2162 TF/03 Include the error message when we fail to unlink a spool file.
2164 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2165 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2166 for maintaining out-of-tree patches for some time.
2168 PP/01 Bugzilla 139: Documentation and portability issues.
2169 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2170 Handle per-OS dynamic-module compilation flags.
2172 PP/02 Let /dev/null have normal permissions.
2173 The 4.73 fixes were a little too stringent and complained about the
2174 permissions on /dev/null. Exempt it from some checks.
2175 Reported by Andreas M. Kirchwitz.
2177 PP/03 Report version information for many libraries, including
2178 Exim version information for dynamically loaded libraries. Created
2179 version.h, now support a version extension string for distributors
2180 who patch heavily. Dynamic module ABI change.
2182 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2183 privilege escalation vulnerability whereby the Exim run-time user
2184 can cause root to append content of the attacker's choosing to
2187 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2190 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2191 If dropping privileges for untrusted macros, we disabled normal logging
2192 on the basis that it would fail; for the Exim run-time user, this is not
2193 the case, and it resulted in successful deliveries going unlogged.
2194 Fixed. Reported by Andreas Metzler.
2200 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2201 only prepend for the Resent-* case. Fixes regression introduced in
2202 Exim 4.70 by NM/22 for Bugzilla 607.
2204 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2205 increasing numbers of administrators be bitten by this.
2207 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2209 PP/03 Bugzilla 994: added openssl_options main configuration option.
2211 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2213 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2215 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2217 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2218 without a peer certificate, leading to a segfault because of an
2219 assumption that peers always have certificates. Be a little more
2220 paranoid. Problem reported by Martin Tscholak.
2222 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2223 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2224 NB: ClamAV planning to remove STREAM in "middle of 2010".
2225 CL also introduces -bmalware, various -d+acl logging additions and
2226 more caution in buffer sizes.
2228 PP/09 Implemented reverse_ip expansion operator.
2230 PP/10 Bugzilla 937: provide a "debug" ACL control.
2232 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2234 PP/12 Bugzilla 973: Implement --version.
2236 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2238 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2240 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2242 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2243 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2244 ignore trailing whitespace.
2246 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2248 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2251 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2252 Notification from Dr Andrew Aitchison.
2254 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2255 ExtendedDetectionInfo response format.
2256 Notification from John Horne.
2258 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2261 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2262 XSL and documented dependency on system catalogs, with examples of how
2265 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2268 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2269 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2270 configuration file which is writeable by the Exim user or group.
2272 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2273 of configuration files to cover files specified with the -C option if
2274 they are going to be used with root privileges, not just the default
2277 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2278 option (effectively making it always true).
2280 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2281 files to be used while preserving root privileges.
2283 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2284 that rogue child processes cannot use them.
2286 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2287 run-time user, instead of root.
2289 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2290 Exim run-time user without dropping privileges.
2292 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2293 result string, instead of calling string_vformat() twice with the same
2296 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2297 for other users. Others should always drop root privileges if they use
2298 -C on the command line, even for a whitelisted configure file.
2300 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2302 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2308 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2309 $data_path, and $header_path variables; fixed documentation bugs and
2312 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2313 exipick to access non-standard spools, including the "frozen" queue
2316 NM/01 Bugzilla 965: Support mysql stored procedures.
2317 Patch from Alain Williams
2319 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2321 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2322 Patch from Andreas Metzler
2324 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2325 Patch from Kirill Miazine
2327 NM/05 Bugzilla 671: Added umask to procmail example.
2329 JJ/03 installed exipick 20100323.0, fixing doc bug
2331 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2332 directory. Notification and patch from Dan Rosenberg.
2334 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2336 TK/02 Improve log output when DKIM signing operation fails.
2338 MH/01 Treat the transport option dkim_domain as a colon separated
2339 list, not as a single string, and sign the message with each element,
2340 omitting multiple occurences of the same signer.
2342 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2343 Bugzilla 985, 986. Patch by Simon Arlott
2345 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2346 Patch by Simon Arlott
2348 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2349 MBX locking. Notification from Dan Rosenberg.
2355 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2357 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2359 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2361 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2363 NM/04 Bugzilla 915: Flag broken perl installation during build.
2369 TK/01 Added patch by Johannes Berg that expands the main option
2370 "spamd_address" if it starts with a dollar sign.
2372 TK/02 Write list of recipients to X-Envelope-Sender header when building
2373 the mbox-format spool file for content scanning (suggested by Jakob
2376 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2377 (http://www.dcc-servers.net/) support via dccifd. Activated by
2378 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2380 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2381 by Mark Daniel Reidel <mr@df.eu>.
2383 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2384 When building exim an external PCRE library is now needed -
2385 PCRE is a system library on the majority of modern systems.
2386 See entry on PCRE_LIBS in EDITME file.
2388 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2389 conversation. Added nologin parameter to request.
2390 Patch contributed by Kirill Miazine.
2392 TF/01 Do not log submission mode rewrites if they do not change the address.
2394 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2396 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2397 log files in place. Contributed by Roberto Lima.
2399 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2401 TF/03 Bugzilla 615: When checking the local_parts router precondition
2402 after a local_part_suffix or local_part_prefix option, Exim now
2403 does not use the address's named list lookup cache, since this
2404 contains cached lookups for the whole local part.
2406 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2407 Robert Millan. Documentation is in experimental-spec.txt.
2409 TF/04 Bugzilla 668: Fix parallel build (make -j).
2411 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2413 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2414 Patch provided by Jan Srzednicki.
2416 TF/05 Leading white space used to be stripped from $spam_report which
2417 wrecked the formatting. Now it is preserved.
2419 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2420 that they are available at delivery time.
2422 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2424 TF/08 TLS error reporting now respects the incoming_interface and
2425 incoming_port log selectors.
2427 TF/09 Produce a more useful error message if an SMTP transport's hosts
2428 setting expands to an empty string.
2430 NM/06 Bugzilla 744: EXPN did not work under TLS.
2431 Patch provided by Phil Pennock.
2433 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2434 Patch provided by Richard Godbee.
2436 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2437 acl_smtp_notquit, added index entry.
2439 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2440 Patch provided by Eugene Bujak.
2442 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2443 accept(). Patch provided by Maxim Dounin.
2445 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2446 Patch provided by Phil Pennock.
2448 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2450 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2451 Patch provided by Brad "anomie" Jorsch.
2453 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2454 Patch provided by Dean Brooks.
2456 TK/05 Add native DKIM support (does not depend on external libraries).
2458 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2459 Patch provided by Graeme Fowler.
2461 NM/16 Bugzilla 851: Documentation example syntax fix.
2463 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2465 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2468 NM/19 Bugzilla 745: TLS version reporting.
2469 Patch provided by Phil Pennock.
2471 NM/20 Bugzilla 167: bool: condition support.
2472 Patch provided by Phil Pennock.
2474 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2475 clients. Patch provided by Phil Pennock.
2477 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2478 Patch provided by Brad "anomie" Jorsch.
2480 NM/23 Bugzilla 687: Fix misparses in eximstats.
2481 Patch provided by Heiko Schlittermann.
2483 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2484 Patch provided by Heiko Schlittermann.
2486 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2487 plus update to original patch.
2489 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2491 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2492 Patch provided by David Brownlee.
2494 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2496 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2498 NM/30 Bugzilla 888: TLS documentation bugfixes.
2500 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2502 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2503 Unlike the original bugzilla I have changed all shell scripts in src tree.
2505 NM/33 Bugzilla 898: Transport filter timeout fix.
2506 Patch by Todd Rinaldo.
2508 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2509 Patch by Serge Demonchaux.
2511 NM/35 Bugzilla 39: Base64 decode bug fixes.
2512 Patch by Jakob Hirsch.
2514 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2516 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2518 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2520 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2526 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2527 ALT-N's libdkim that I have put here:
2528 http://duncanthrax.net/exim-experimental/
2530 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2531 were called 'true' and it seems that C99 defines that as a reserved
2532 keyword to be used with 'bool' variable types. That means you could
2533 not include C99-style headers which use bools without triggering
2534 build errors in sieve.c.
2536 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2537 as mailq or other aliases. Changed the --help handling significantly
2538 to do whats expected. exim_usage() emits usage/help information.
2540 SC/01 Added the -bylocaldomain option to eximstats.
2542 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2544 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2546 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2552 PH/01 Another patch from the Sieve maintainer.
2554 PH/02 When an IPv6 address is converted to a string for single-key lookup
2555 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2556 dots are used instead of colons so that keys in lsearch files need not
2557 contain colons. This was done some time before quoting was made available
2558 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2559 (notated using the quote facility) so as to distinguish them from IPv4
2560 keys. This meant that lookups for IP addresses in host lists did not work
2561 for iplsearch lookups.
2563 This has been fixed by arranging for IPv6 addresses to be expressed with
2564 colons if the lookup type is iplsearch. This is not incompatible, because
2565 previously such lookups could never work.
2567 The situation is now rather anomalous, since one *can* have colons in
2568 ordinary lsearch keys. However, making the change in all cases is
2569 incompatible and would probably break a number of configurations.
2571 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2574 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2575 conversion specification without a maximum field width, thereby enabling
2576 a rogue spamd server to cause a buffer overflow. While nobody in their
2577 right mind would setup Exim to query an untrusted spamd server, an
2578 attacker that gains access to a server running spamd could potentially
2579 exploit this vulnerability to run arbitrary code as the Exim user.
2581 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2582 $primary_hostname instead of what libspf2 thinks the hosts name is.
2584 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2585 a directory entry by the name of the lookup key. Previously, if a
2586 symlink pointed to a non-existing file or a file in a directory that
2587 Exim lacked permissions to read, a lookup for a key matching that
2588 symlink would fail. Now it is enough that a matching directory entry
2589 exists, symlink or not. (Bugzilla 503.)
2591 PH/03 The body_linecount and body_zerocount variables are now exported in the
2594 PH/04 Added the $dnslist_matched variable.
2596 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2597 This means they are set thereafter only if the connection becomes
2600 PH/06 Added the client_condition to authenticators so that some can be skipped
2601 by clients under certain conditions.
2603 PH/07 The error message for a badly-placed control=no_multiline_responses left
2604 "_responses" off the end of the name.
2606 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2608 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2609 (without spaces) instead of just copying the configuration text.
2611 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2613 PH/11 Added $max_received_linelength.
2615 PH/12 Added +ignore_defer and +include_defer to host lists.
2617 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2618 way in which PCRE > 7.0 is built.
2620 PH/14 Implemented queue_only_load_latch.
2622 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2623 MAIL command. The effect was to mangle the value on 64-bit systems.
2625 PH/16 Another patch from the Sieve maintainer.
2627 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2629 PH/18 If a system quota error occurred while trying to create the file for
2630 a maildir delivery, the message "Mailbox is full" was not appended to the
2631 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2632 only to a quota excession during the actual writing of the file.
2634 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2635 characters?) which causes problems in log lines. The DN values are now
2636 passed through string_printing() before being added to log lines.
2638 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2639 and InterBase are left for another time.)
2641 PH/21 Added message_body_newlines option.
2643 PH/22 Guard against possible overflow in moan_check_errorcopy().
2645 PH/23 POSIX allows open() to be a macro; guard against that.
2647 PH/24 If the recipient of an error message contained an @ in the local part
2648 (suitably quoted, of course), incorrect values were put in $domain and
2649 $local_part during the evaluation of errors_copy.
2655 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2656 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2659 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2660 issue a MAIL command.
2662 PH/02 In an ACL statement such as
2664 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2666 if a client was not listed at all, or was listed with a value other than
2667 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2668 the condition was not true (as it should be), so access was not denied.
2669 The bug was that the ! inversion was incorrectly passed on to the second
2670 item. This has been fixed.
2672 PH/03 Added additional dnslists conditions == and =& which are different from
2673 = and & when the dns lookup returns more than one IP address.
2675 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2676 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2678 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2679 FSYNC, which compiles an option called disable_fsync that allows for
2680 bypassing fsync(). The documentation is heavily laced with warnings.
2682 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2684 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2685 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2686 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2687 including adding "make clean"; (3) Added -fPIC when compiling the test
2688 dynamically loaded module, to get rid of a warning.
2690 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2691 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2692 = 0s. The bug is otherwise harmless.
2694 PH/07 There was a bug in the dovecot authenticator such that the value of
2695 $auth1 could be overwritten, and so not correctly preserved, after a
2696 successful authentication. This usually meant that the value preserved by
2697 the server_setid option was incorrect.
2699 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2701 PH/09 Installed PCRE release 7.0.
2703 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2704 run for batched SMTP input. It is now run at the start of every message
2705 in the batch. While fixing this I discovered that the process information
2706 (output by running exiwhat) was not always getting set for -bs and -bS
2707 input. This is fixed, and it now also says "batched" for BSMTP.
2709 PH/11 Added control=no_pipelining.
2711 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2712 patch, slightly modified), and move the expansion of helo_data till after
2713 the connection is made in the smtp transport (so it can use these
2716 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2718 PH/14 Added log_selector = +pid.
2720 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2722 PH/16 Add ${if forany and ${if forall.
2724 PH/17 Added dsn_from option to vary the From: line in DSNs.
2726 PH/18 Flush SMTP output before performing a callout, unless control =
2727 no_callout_flush is set.
2729 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2730 was true (the default) a successful delivery failed to delete the retry
2731 item, thus causing premature timeout of the address. The bug is now
2734 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2736 PH/21 Long custom messages for fakedefer and fakereject are now split up
2737 into multiline responses in the same way that messages for "deny" and
2738 other ACL rejections are.
2740 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2741 with slight modification.
2743 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2744 draft, changing the syntax and factoring some duplicate code".
2746 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2747 for deliveries of the second and subsequent messages over the same SMTP
2750 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2751 ${reduce, with only minor "tidies".
2753 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2755 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2756 expansion side effects.
2758 PH/27 When a message times out after an over-quota error from an Exim-imposed
2759 quota, the bounce message says "mailbox is full". This message was not
2760 being given when it was a system quota that was exceeded. It now should
2763 MH/03 Made $recipients available in local_scan(). local_scan() already has
2764 better access to the recipient list through recipients_list[], but
2765 $recipients can be useful in postmaster-provided expansion strings.
2767 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2768 in the case of a MAIL command with additional options following the
2769 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2770 were accidentally chopped off.
2772 PH/29 SMTP synchronization checks are implemented when a command is read -
2773 there is a check that no more input is waiting when there shouldn't be
2774 any. However, for some commands, a delay in an ACL can mean that it is
2775 some time before the response is written. In this time, more input might
2776 arrive, invalidly. So now there are extra checks after an ACL has run for
2777 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2778 pipelining has not been advertised.
2780 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2782 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2783 returned in the SMTP response when smtp_return_error_details was false.
2784 This has been fixed.
2786 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2787 instead of the C I/O that was originally supplied, because problems were
2788 reported on Solaris.
2790 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2791 Exim which did not show up earlier: it was assuming that a call to
2792 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2793 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2794 was a macro that became an assignment, so it seemed to work. This has
2795 changed to a proper function call with a void return, hence the compile
2796 error. Exim's code has been fixed.
2798 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2801 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2803 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2805 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2806 because it turns out to be too big or there is a timeout) there is no
2807 "Completed" line in the log. When some messages of this type were
2808 selected by exigrep, they were listed as "not completed". Others were
2809 picked up by some special patterns. I have improved the selection
2810 criteria to be more general.
2812 PH/38 The host_find_failed option in the manualroute router can now be set
2813 to "ignore", to completely ignore a host whose IP address cannot be
2814 found. If all hosts are ignored, the behaviour is controlled by the new
2815 host_all_ignored option.
2817 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2818 homing or because of multiple MX records with /mx) generated more than
2819 one IP address, and the following item turned out to be the local host,
2820 all the secondary addresses of the first item were incorrectly removed
2821 from the list, along with the local host and any following hosts (which
2822 is what is supposed to happen).
2824 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2825 whoever called Exim into the -H file. In the case of the daemon it was
2826 behaving confusingly. When first started, it used values for whoever
2827 started the daemon, but after a SIGHUP it used the Exim user (because it
2828 calls itself on a restart). I have changed the code so that it now always
2831 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2832 message are rejected with the same error (e.g. no authentication or bad
2833 sender address), and a DATA command is nevertheless sent (as can happen
2834 with PIPELINING or a stupid MUA), the error message that was given to the
2835 RCPT commands is included in the rejection of the DATA command. This is
2836 intended to be helpful for MUAs that show only the final error to their
2839 PH/42 Another patch from the Sieve maintainer.
2841 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2842 Eximstats - Fixed some broken HTML links and added missing column headers
2844 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2845 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2847 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2849 PH/43 Yet another patch from the Sieve maintainer.
2851 PH/44 I found a way to check for a TCP/IP connection going away before sending
2852 the response to the final '.' that terminates a message, but only in the
2853 case where the client has not sent further data following the '.'
2854 (unfortunately, this is allowed). However, in many cases there won't be
2855 any further data because there won't be any more messages to send. A call
2856 to select() can be used: if it shows that the input is "ready", there is
2857 either input waiting, or the socket has been closed. An attempt to read
2858 the next input character can distinguish the two cases. Previously, Exim
2859 would have sent an OK response which the client would never have see.
2860 This could lead to message repetition. This fix should cure that, at
2861 least in a lot of common cases.
2863 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2864 advertised in response to EHLO.
2870 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2871 fixed by 4.65/MH/01 (is this a record?) are fixed:
2873 (i) An empty string was always treated as zero by the numeric comparison
2874 operators. This behaviour has been restored.
2876 (ii) It is documented that the numeric comparison operators always treat
2877 their arguments as decimal numbers. This was broken in that numbers
2878 starting with 0 were being interpreted as octal.
2880 While fixing these problems I realized that there was another issue that
2881 hadn't been noticed. Values of message_size_limit (both the global option
2882 and the transport option) were treated as octal if they started with 0.
2883 The documentation was vague. These values are now always treated as
2884 decimal, and I will make that clear in the documentation.
2890 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2891 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2894 MH/01 Don't check that the operands of numeric comparison operators are
2895 integers when their expansion is in "skipping" mode (fixes bug
2896 introduced by 4.64-PH/07).
2898 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2899 child addresses, Exim now panics and dies. Previously, because the count
2900 is held in a short int, deliveries were likely to be lost. As such a
2901 large number of recipients for a single message is ridiculous
2902 (performance will be very, very poor), I have chosen to impose a limit
2903 rather than extend the field.
2909 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2910 leftover -K file (the existence of which was triggered by #402).
2911 While we were at it, introduced process PID as part of the -K
2912 filename. This should rule out race conditions when creating
2915 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2916 processing considerably. Previous code took too long for large mails,
2917 triggering a timeout which in turn triggers #401.
2919 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2920 in the DK code in transports.c. sendfile() is not really portable,
2921 hence the _LINUX specificness.
2923 TF/01 In the add_headers option to the mail command in an Exim filter,
2924 there was a bug that Exim would claim a syntax error in any
2925 header after the first one which had an odd number of characters
2928 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2929 callout verification, Exim cached a "reject" for the entire domain. This
2930 is correct for most verifications, but it is not correct for a recipient
2931 verification with use_sender or use_postmaster set, because in that case
2932 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2933 case of MAIL FROM:<> rejection from other early rejections (e.g.
2934 rejection of HELO). When verifying a recipient using a non-null MAIL
2935 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2936 Whatever the result of the callout, the value of the domain cache is
2937 left unchanged (for any other kind of callout, getting as far as trying
2938 RCPT means that the domain itself is ok).
2940 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2943 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2944 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2945 session, the daemon ignores ECONNECT errors and logs others; it now
2946 ignores EPIPE as well.
2948 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2949 (quoted-printable decoding).
2951 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2952 later the small subsequent patch to fix an introduced bug.
2954 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2956 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2958 PH/08 An error is now given if message_size_limit is specified negative.
2960 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2961 to be given (somewhat) arbitrary names.
2963 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2966 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2967 miscellaneous code fixes
2969 PH/10 Added the log_reject_target ACL modifier to specify where to log
2972 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2973 hostname. This is wrong, because it relates to the incoming message (and
2974 probably the interface on which it is arriving) and not to the outgoing
2975 callout (which could be using a different interface). This has been
2976 changed to use the value of the helo_data option from the smtp transport
2977 instead - this is what is used when a message is actually being sent. If
2978 there is no remote transport (possible with a router that sets up host
2979 addresses), $smtp_active_hostname is used.
2981 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2982 tweaks were necessary in order to get it to work (see also 21 below):
2983 (a) The code assumed that strncpy() returns a negative number on buffer
2984 overflow, which isn't the case. Replaced with Exim's string_format()
2986 (b) There were several signed/unsigned issues. I just did the minimum
2987 hacking in of casts. There is scope for a larger refactoring.
2988 (c) The code used strcasecmp() which is not a standard C function.
2989 Replaced with Exim's strcmpic() function.
2990 (d) The code set only $1; it now sets $auth1 as well.
2991 (e) A simple test gave the error "authentication client didn't specify
2992 service in request". It would seem that Dovecot has changed its
2993 interface. Fortunately there's a specification; I followed it and
2994 changed what the client sends and it appears to be working now.
2996 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2999 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3000 address A is aliased to B and C, where B exists and C does not. Without
3001 -v the output is "A verified" because verification stops after a
3002 successful redirection if more than one address is generated. However,
3003 with -v the child addresses are also verified. Exim was outputting "A
3004 failed to verify" and then showing the successful verification for C,
3005 with its parentage. It now outputs "B failed to verify", showing B's
3006 parentage before showing the successful verification of C.
3008 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3009 look up a TXT record in a specific list after matching in a combined
3012 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3013 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3014 they consult the DNS. I had assumed they would set it the way they
3015 wanted; and indeed my experiments on Linux seem to show that in some
3016 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3017 To be on the safe side, however, I have now made the interface to
3018 host_find_byname() similar to host_find_bydns(), with an argument
3019 containing the DNS resolver options. The host_find_byname() function now
3020 sets these options at its start, just as host_find_bydns() does. The smtp
3021 transport options dns_qualify_single and dns_search_parents are passed to
3022 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3023 of host_find_byname() use the default settings of RES_DEFNAMES
3024 (qualify_single) but not RES_DNSRCH (search_parents).
3026 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3027 spool_read_header() do less string testing, by means of a preliminary
3028 switch on the second character of optional "-foo" lines. (This is
3029 overdue, caused by the large number of possibilities that now exist.
3030 Originally there were few.) While I was there, I also converted the
3031 str(n)cmp tests so they don't re-test the leading "-" and the first
3032 character, in the hope this might squeeze out yet more improvement.
3034 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3035 flag allowing group syntax was set by the header_syntax check but not
3036 turned off, possible causing trouble later; (2) The flag was not being
3037 set at all for the header_verify test, causing "group"-style headers to
3038 be rejected. I have now set it in this case, and also caused header_
3039 verify to ignore an empty address taken from a group. While doing this, I
3040 came across some other cases where the code for allowing group syntax
3041 while scanning a header line wasn't quite right (mostly, not resetting
3042 the flag correctly in the right place). These bugs could have caused
3043 trouble for malformed header lines. I hope it is now all correct.
3045 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3046 with the "reply" argument non-NULL. The code, however (which originally
3047 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3048 but it didn't always do it. This confused somebody who was copying the
3049 code for some other use. I have removed all the tests.
3051 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3052 feature that was used to support insecure browsers during the U.S. crypto
3053 embargo. It requires special client support, and Exim is probably the
3054 only MTA that supported it -- and would never use it because real RSA is
3055 always available. This code has been removed, because it had the bad
3056 effect of slowing Exim down by computing (never used) parameters for the
3057 RSA_EXPORT functionality.
3059 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3060 authenticator to fail if there's a tab character in the incoming data
3061 (there should never be unless someone is messing about, as it's supposed
3062 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3063 if the connection is using TLS or if the remote IP is the same as the
3064 local IP, and the "valid-client-cert option" if a client certificate has
3067 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3068 authenticators. This can be used for authorization after authentication
3069 succeeds. (In the case of plaintext, it servers for both authentication
3072 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3073 if any retry times were supplied.
3075 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3076 connection, where there is no client IP address to check. In this
3077 situation, the verify now always succeeds.
3079 PH/25 Applied John Jetmore's -Mset patch.
3081 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3083 PH/27 In a string expansion for a processed (not raw) header when multiple
3084 headers of the same name were present, leading whitespace was being
3085 removed from all of them, but trailing whitespace was being removed only
3086 from the last one. Now trailing whitespace is removed from each header
3087 before concatenation. Completely empty headers in a concatenation (as
3088 before) are ignored.
3090 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3091 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3093 PH/29 [Removed. This was a change that I later backed out, and forgot to
3094 correct the ChangeLog entry (that I had efficiently created) before
3095 committing the later change.]
3097 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3098 address errors (4xx response to RCPT) over the same connection as other
3099 messages routed to the same hosts. Such deliveries are always "forced",
3100 so retry times are not inspected. This resulted in far too many retries
3101 for the affected addresses. The effect occurred only when there were more
3102 hosts than the hosts_max_try setting in the smtp transport when it had
3103 the 4xx errors. Those hosts that it had tried were not added to the list
3104 of hosts for which the message was waiting, so if all were tried, there
3105 was no problem. Two fixes have been applied:
3107 (i) If there are any address or message errors in an SMTP delivery, none
3108 of the hosts (tried or untried) are now added to the list of hosts
3109 for which the message is waiting, so the message should not be a
3110 candidate for sending over the same connection that was used for a
3111 successful delivery of some other message. This seems entirely
3112 reasonable: after all the message is NOT "waiting for some host".
3113 This is so "obvious" that I'm not sure why it wasn't done
3114 previously. Hope I haven't missed anything, but it can't do any
3115 harm, as the worst effect is to miss an optimization.
3117 (ii) If, despite (i), such a delivery is accidentally attempted, the
3118 routing retry time is respected, so at least it doesn't keep
3119 hammering the server.
3121 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3122 in ${readsocket because some servers need this prod.
3124 PH/32 Added some extra debug output when updating a wait-xxx database.
3126 PH/33 The hint "could be header name not terminated by colon", which has been
3127 given for certain expansion errors for a long time, was not being given
3128 for the ${if def:h_colon_omitted{... case.
3130 PH/34 The spec says: "With one important exception, whenever a domain list is
3131 being scanned, $domain contains the subject domain." There was at least
3132 one case where this was not true.
3134 PH/35 The error "getsockname() failed: connection reset by peer" was being
3135 written to the panic log as well as the main log, but it isn't really
3136 panic-worthy as it just means the connection died rather early on. I have
3137 removed the panic log writing for the ECONNRESET error when getsockname()
3140 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3141 runs only) independently of the message's sender address. This meant
3142 that, if the 4xx error was in fact related to the sender, a different
3143 message to the same recipient with a different sender could confuse
3144 things. In particular, this can happen when sending to a greylisting
3145 server, but other circumstances could also provoke similar problems.
3146 I have changed the default so that the retry time for these errors is now
3147 based a combination of the sender and recipient addresses. This change
3148 can be overridden by setting address_retry_include_sender=false in the
3151 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3152 remote server are returned as part of bounce messages. This was not
3153 happening for LMTP over a pipe (the lmtp transport), but now it is the
3154 same for both kinds of LMTP.
3156 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3157 in header lines that were in fact CNAMEs. This is no longer the case.
3159 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3160 and queue runs started by the daemon processed all messages. This has
3161 been fixed so that -R and -S can now usefully be given with -q<time>.
3163 PH/40 Import PCRE release 6.7 (fixes some bugs).
3165 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3167 PH/42 Give an error if -q is specified more than once.
3169 PH/43 Renamed the variables $interface_address and $interface_port as
3170 $received_ip_address and $received_port, to make it clear that these
3171 values apply to message reception, and not to the outgoing interface when
3172 a message is delivered. (The old names remain recognized, of course.)
3174 PH/44 There was no timeout on the connect() call when using a Unix domain
3175 socket in the ${readsocket expansion. There now is.
3177 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3178 be meaningful with "accept".
3180 SC/01 Eximstats V1.43
3181 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3183 SC/02 Eximstats V1.44
3184 Use a glob alias rather than an array ref in the generated
3185 parser. This improves both readability and performance.
3187 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3188 Collect SpamAssassin and rejection statistics.
3189 Don't display local sender or destination tables unless
3190 there is data to show.
3191 Added average volumes into the top table text output.
3193 SC/04 Eximstats V1.46
3194 Collect data on the number of addresses (recipients)
3195 as well as the number of messages.
3197 SC/05 Eximstats V1.47
3198 Added 'Message too big' to the list of mail rejection
3199 reasons (thanks to Marco Gaiarin).
3201 SC/06 Eximstats V1.48
3202 Mainlog lines which have GMT offsets and are too short to
3203 have a flag are now skipped.
3205 SC/07 Eximstats V1.49 (Alain Williams)
3206 Added the -emptyok flag.
3208 SC/08 Eximstats V1.50
3209 Fixes for obtaining the IP address from reject messages.
3211 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3212 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3213 whitespace changes from 4.64-PH/27
3215 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3218 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3219 are found, allow negative numbers in numeric criteria)
3221 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3223 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3224 to match changes made in 4.64-PH/43
3226 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3228 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3229 "domain" is actually an IP address, and force a failure. This locks out
3230 those revolvers/nameservers that support "A-for-A" lookups, in
3231 contravention of the specifications.
3233 PH/48 When a host name was looked up from an IP address, and the subsequent
3234 forward lookup of the name timed out, the host name was left in
3235 $sender_host_name, contrary to the specification.
3237 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3238 restricted to single-key lookups, Exim was not diagnosing an error if
3239 * or *@ was used with a query-style lookup.
3241 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3243 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3244 long ago, but noone interested enough thought of it. Let's just say that
3245 the "1.1" means that there are some new functions that weren't there at
3246 some point in the past.
3248 PH/51 Error processing for expansion failure of helo_data from an smtp
3249 transport during callout processing was broken.
3251 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3252 tested/used via the -bh/-bhc/-bs options.
3254 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3255 bug, fixed in subsequent PCRE releases).
3257 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3258 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3260 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3266 SC/01 Use a glob alias rather than an array ref in eximstats generated
3267 parser. This improves both readability and performance.
3269 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3270 Don't display local sender or destination tables in eximstats unless
3271 there is data to show.
3272 Added average volumes into the eximstats top table text output.
3274 SC/03 Collect data on the number of addresses (recipients) as well
3275 as the number of messages in eximstats.
3277 TF/01 Correct an error in the documentation for the redirect router. Exim
3278 does (usually) call initgroups() when daemonizing.
3280 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3281 with consistent privilege compared to when running as a daemon.
3283 TF/03 Note in the spec that $authenticated_id is not set for local
3284 submissions from trusted users.
3286 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3287 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3289 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3290 by adding some example configuration directives to the default
3291 configuration file. A little bit of work is required to uncomment the
3292 directives and define how usernames and passwords are checked, but
3293 there is now a framework to start from.
3295 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3296 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3297 without this. I don't know how relevant this is to other LDAP libraries.
3299 PH/02 Add the verb name to the "unknown ACL verb" error.
3301 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3303 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3305 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3306 directory not expanded when it should be if an expanded home directory
3307 was set for the address (which is overridden by the transport).
3309 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3312 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3313 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3314 because it is too late at that time, and has no effect.
3316 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3317 security issue with \' (bugzilla #107). I could not use the
3318 PQescapeStringConn() function, because it needs a PGconn value as one of
3321 PH/08 When testing addresses using -bt, indicate those final addresses that
3322 are duplicates that would not cause an additional delivery. At least one
3323 person was confused, thinking that -bt output corresponded to deliveries.
3324 (Suppressing duplicates isn't a good idea as you lose the information
3325 about possibly different redirections that led to the duplicates.)
3327 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3328 systems where poll() doesn't work, in particular OS X.
3330 PH/10 Added more information to debugging output for retry time not reached.
3332 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3333 operations in malware.c.
3335 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3338 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3339 syslog_duplication set false, log lines that would normally be written
3340 both the the main log and to the reject log were not written to syslog at
3343 PH/14 In the default configuration, change the use of "message" in ACL warn
3344 statements to "add_header".
3346 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3347 not followed by a command (e.g. "seen endif").
3349 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3350 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3353 PH/17 Added extra conditions to the default value of delay_warning_condition
3357 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3358 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3359 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3362 The Auto-Submitted: and various List- headers are standardised, whereas I
3363 don't think Precedence: ever was.
3365 PH/18 Refactored debugging code in route_finduser() to show more information,
3366 in particular, the error code if getpwnam() issues one.
3368 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3369 This is apparently needed in addition to the PH/07 change above to avoid
3370 any possible encoding problems.
3372 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3373 but not after initializing Perl.
3375 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3376 output them only if debugging. By default they are written stderr,
3377 apparently, which is not desirable.
3379 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3382 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3385 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3387 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3388 authenticated or an ident call has been made. Suppress the default
3389 values for $authenticated_id and $authenticated_sender (but permit -oMai
3390 and -oMas) when testing with -bh.
3392 PH/24 Re-jigged the order of the tests in the default configuration so that the
3393 tests for valid domains and recipients precede the DNS black list and CSA
3394 tests, on the grounds that those ones are more expensive.
3396 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3397 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3398 This bug exists in every version of Exim that I still have, right back to
3401 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3402 However, an attempt to turn on case-sensitivity in a regex key by
3403 including (?-i) didn't work because the subject string was already
3404 lowercased, and the effects were non-intuitive. It turns out that a
3405 one-line patch can be used to allow (?-i) to work as expected.
3411 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3412 other effects) broke the use of negated acl sub-conditions.
3414 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3417 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3418 "Deny" causes Exim to reject the incoming connection with a 554 error.
3419 Unfortunately, if there is a major crisis, such as a disk failure,
3420 tcp-wrappers gives "deny", whereas what one would like would be some
3421 kind of temporary error. A kludge has been added to help with this.
3422 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3423 554 error is used if errno is still zero or contains ENOENT (which occurs
3424 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3427 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3429 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3430 errors. Otherwise a message that provokes a temporary error (when other
3431 messages do not) can cause a whole host to time out.
3433 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3434 addresses were routed directly to files or pipes from a redirect router.
3435 File deliveries just didn't batch; pipe deliveries might have suffered
3438 PH/06 A failure to get a lock for a hints database would erroneously always say
3439 "Failed to get write lock", even when it was really a read lock.
3441 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3442 of 0600. This has been changed to use the value of the lockfile_mode
3443 option (which defaults to 0600).
3445 PH/08 Applied small patch from the Sieve maintainer.
3447 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3448 folder from quota calculations, a direct delivery into this folder messed
3449 up the contents of the maildirsize file. This was because the regex was
3450 used only to exclude .Trash (or whatever) when the size of the mailbox
3451 was calculated. There was no check that a delivery was happening into an
3452 excluded directory. This bug has been fixed by ignoring all quota
3453 processing for deliveries into excluded directories.
3455 PH/10 Added the maildirfolder_create_regex option to appendfile.
3461 PH/01 The code for finding all the local interface addresses on a FreeBSD
3462 system running IPv6 was broken. This may well have applied to all BSD
3463 systems, as well as to others that have similar system calls. The broken
3464 code found IPv4 interfaces correctly, but gave incorrect values for the
3465 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3466 that it would not match correctly against @[] and not recognize the IPv6
3469 PH/02 The ipliteral router was not recognizing addresses of the form user@
3470 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3472 PH/03 Added disable_ipv6.
3474 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3475 decoded form, because it is most often used to construct To: headers
3476 lines in autoreplies, and the decoded form may well be syntactically
3477 invalid. However, $reply_address has leading white space removed, and all
3478 newlines turned into spaces so that the autoreply transport does not
3481 PH/05 If group was specified without a user on a router, and no group or user
3482 was specified on a transport, the group from the router was ignored.
3484 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3485 for visible compile-time settings that can be used to change these
3486 numbers, for those that want even more. Backwards compatibility with old
3487 spool files has been maintained. However, going back to a previous Exim
3488 release will lost any variables that are in spool files.
3490 PH/07 Two small changes when running in the test harness: increase delay when
3491 passing a TCP/IP connection to a new process, in case the original
3492 process has to generate a bounce, and remove special handling of
3493 127.0.0.2 (sic), which is no longer necessary.
3495 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3496 be the same on different OS.
3498 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3501 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3502 whether --show-vars was specified or not
3504 JJ/02 exipick: Added support for new ACL variable spool format introduced
3507 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3508 syntactically invalid From: or Reply-to: line, and a filter used this to
3509 generate an autoreply, and therefore failed to obtain an address for the
3510 autoreply, Exim could try to deliver to a non-existent relative file
3511 name, causing unrelated and misleading errors. What now happens is that
3512 it logs this as a hard delivery error, but does not attempt to create a
3515 PH/11 The exinext utility has a -C option for testing purposes, but although
3516 the given file was scanned by exinext itself; it wasn't being passed on
3517 when Exim was called.
3519 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3520 an end-of-file indication when reading a command response.
3522 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3523 compiled. In many other places in Exim, IPv6 addresses are always
3524 recognized, so I have changed this. It also means that IPv4 domain
3525 literals of the form [IPV4:n.n.n.n] are now always recognized.
3527 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3528 used if the router is not running as root, for example, when verifying at
3529 ACL time, or when using -bh. The debugging output from this situation was
3530 non-existent - all you got was a failure to exec. I have made two
3533 (a) Failures to set uid/gid, the current directory, or a process leader
3534 in a subprocess such as that created by queryprogram now generate
3535 suitable debugging output when -d is set.
3537 (b) The queryprogram router detects when it is not running as root,
3538 outputs suitable debugging information if -d is set, and then runs
3539 the subprocess without attempting to change uid/gid.
3541 PH/15 Minor change to Makefile for building test_host (undocumented testing
3544 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3545 additional section of a DNS packet that returns MX or SRV records.
3546 Instead, it always explicitly searches for A/AAAA records. This avoids
3547 major problems that occur when a DNS server includes only records of one
3548 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3549 fixed another bug: if SRV records were looked up and the corresponding
3550 address records were *not* found in the additional section, the port
3551 values from the SRV records were lost.
3553 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3554 using the correct key (the original address) when searching the retry
3555 rules in order to find which one to use for generating the retry hint.
3557 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3558 adding the default one. Similarly, if it contains a Reply-To: header, the
3559 errors_reply_to option, if set, is not used.
3561 PH/19 When calculating a retry time, Exim used to measure the "time since
3562 failure" by looking at the "first failed" field in the retry record. Now
3563 it does not use this if it is later than than the arrival time of the
3564 message. Instead it uses the arrival time. This makes for better
3565 behaviour in cases where some deliveries succeed, thus re-setting the
3566 "first failed" field. An example is a quota failure for a huge message
3567 when small messages continue to be delivered. Without this change, the
3568 "time since failure" will always be short, possible causing more frequent
3569 delivery attempts for the huge message than are intended.
3570 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3572 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3573 $1, $2, $3) because the numerical variables can be reset during some
3574 expansion items (e.g. "match"), thereby losing the authentication data.
3576 PH/21 Make -bV show the size of off_t variables so that the test suite can
3577 decide whether to run tests for quotas > 2G.
3579 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3580 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3581 filecount value is greater than 2G or if a quota value is greater than 2G
3582 on a system where the size of off_t is not greater than 4, a panic error
3585 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3586 never match. The debug and -bh output now contains an explicit error
3587 message indicating a malformed IPv4 address or mask.
3589 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3590 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3591 PH/23 above applies.
3593 PH/25 Do not write to syslog when running in the test harness. The only
3594 occasion when this arises is a failure to open the main or panic logs
3595 (for which there is an explicit test).
3597 PH/26 Added the /no_tell option to "control=freeze".
3599 PH/27 If a host name lookup failed very early in a connection, for example, if
3600 the IP address matched host_lookup and the reverse lookup yielded a name
3601 that did not have a forward lookup, an error message of the form "no IP
3602 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3603 could be logged. Now it outputs the IP address instead of "NULL".
3605 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3606 allows the sender and the authenticated sender to be set when
3607 submitting a message from within Exim. Since child_open_exim() is
3608 documented for local_scan(), the new function should be too.
3610 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3611 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3612 results in an empty string is now treated as unset.
3614 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3616 PH/31 Added sender_verify_fail logging option.
3618 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3619 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3620 tidied the source and removed it altogether.
3622 PH/33 When a queue run was abandoned because the load average was too high, a
3623 log line was always written; now it is written only if the queue_run log
3624 selector is set. In addition, the log line for abandonment now contains
3625 information about the queue run such as the pid. This is always present
3626 in "start" and "stop" lines but was omitted from the "abandon" line.
3628 PH/34 Omit spaces between a header name and the colon in the error message that
3629 is given when verify = headers_syntax fails (if there are lots of them,
3630 the message gets confusing).
3632 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3633 names, as there are now some PTR records that contain slashes. This check
3634 is only to protect against broken name servers that fall over on strange
3635 characters, so the fact that it applies to all lookups doesn't matter.
3637 PH/36 Now that the new test suite is complete, we can remove some of the
3638 special code in Exim that was needed for the old test suite. For example,
3639 sorting DNS records because real resolvers return them in an arbitrary
3640 order. The new test suite's fake resolver always returns records in the
3643 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3644 bounces) except when queue_only is set, to avoid logging races between
3645 the different processes.
3647 PH/38 Panic-die if .include specifies a non-absolute path.
3649 PH/39 A tweak to the "H" retry rule from its user.
3651 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3652 a label. They prevented compilation on older perls.
3654 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3655 a warning to be raised on newish perls.
3657 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3658 on queue. Changes to match documented behaviour of showing count of
3659 messages matching specified criteria.
3661 PH/40 Changed the default ident timeout from 30s to 5s.
3663 PH/41 Added support for the use of login_cap features, on those BSD systems
3664 that have them, for controlling the resources used by pipe deliveries.
3666 PH/42 The content-scanning code uses fopen() to create files in which to put
3667 message data. Previously it was not paying any attention to the mode of
3668 the files. Exim runs with umask(0) because the rest of the code creates
3669 files with open(), and sets the required mode explicitly. Thus, these
3670 files were ending up world-writeable. This was not a big issue, because,
3671 being within the spool directory, they were not world-accessible. I have
3672 created a function called modefopen, which takes an additional mode
3673 argument. It sets umask(777), creates the file, chmods it to the required
3674 mode, then resets the umask. All the relevant calls to fopen() in the
3675 content scanning code have been changed to use this function.
3677 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3678 to 24 hours. This avoids potential overflow problems when processing G
3679 and H retry rules. I suspect nobody ever tinkers with this value.
3681 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3683 PH/45 When the plaintext authenticator is running as a client, the server's
3684 challenges are checked to ensure they are valid base64 strings. By
3685 default, the authentication attempt is cancelled if an invalid string is
3686 received. Setting client_ignore_invalid_base64 true ignores these errors.
3687 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3688 they are received. Thus, the responses can be made to depend on the
3689 challenges. If an invalid string is ignored, an empty string is placed in
3692 PH/46 Messages that are created by the autoreply transport now contains a
3693 References: header, in accordance with RFCs 2822 and 3834.
3695 PH/47 Added authenticated_sender_force to the smtp transport.
3697 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3699 PH/49 Installed latest patch from the Sieve maintainer.
3701 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3702 was also set, the appendfile transport was unnecessarily scanning a
3703 directory of message files (e.g. for maildir delivery) to find the count
3704 of files (along with the size), even though it did not need this
3705 information. It now does the scan only if it needs to find either the
3706 size of the count of files.
3708 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3710 PH/52 Two bugs concerned with error handling when the smtp transport is
3713 (i) Exim was not creating retry information for temporary errors given
3714 for individual recipients after the DATA command when the smtp transport
3715 was used in LMTP mode. This meant that they could be retried too
3716 frequently, and not timed out correctly.
3718 (ii) Exim was setting the flag that allows error details to be returned
3719 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3720 recipients that were returned after the DATA command.
3722 PH/53 This is related to PH/52, but is more general: for any failing address,
3723 when detailed error information was permitted to be returned to the
3724 sender, but the error was temporary, then after the final timeout, only
3725 "retry timeout exceeded" was returned. Now it returns the full error as
3726 well as "retry timeout exceeded".
3728 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3729 do this, and (what is worse) MTAs that accept it.
3731 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3732 will now be deprecated.
3734 PH/56 New os.c-cygwin from the Cygwin maintainer.
3736 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3737 formats (previously only available in exim formats via -bpr, -bpru,
3738 and -bpra. Now also available in native and exiqgrep formats)
3740 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3741 with very large, slow to parse queues
3743 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3745 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3747 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3748 responses to authentication challenges, though it was showing the
3749 challenges; (ii) I've removed the CR characters from the debug output for
3752 PH/58 Allow for the insertion of a newline as well as a space when a string
3753 is turned into more than one encoded-word during RFC 2047 encoding. The
3754 Sieve code now uses this.
3756 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3757 data_4xx, lost_connection, tls_required.
3759 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3760 message was being sent as an SMTP response.
3762 PH/61 Add -l and -k options to exicyclog.
3764 PH/62 When verifying, if an address was redirected to one new address, so that
3765 verification continued, and the new address failed or deferred after
3766 having set something in $address_data, the value of $address_data was not
3767 passed back to the ACL. This was different to the case when no
3768 redirection occurred. The value is now passed back in both cases.
3770 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3771 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3772 use login_cap.h, so on its own it isn't the distinguishing feature. The
3773 new name refers directly to the setclassresources() function.
3775 PH/65 Added configuration files for NetBSD3.
3777 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3779 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3780 is preferred over IPv4.
3782 PH/68 The bounce_return_message and bounce_return_body options were not being
3783 honoured for bounces generated during the reception of non-SMTP messages.
3784 In particular, this applied to messages rejected by the ACL. This bug has
3785 been fixed. However, if bounce_return_message is true and bounce_return_
3786 body is false, the headers that are returned for a non-SMTP message
3787 include only those that have been read before the error was detected.
3788 (In the case of an ACL rejection, they have all been read.)
3790 PH/69 The HTML version of the specification is now built in a directory called
3791 spec_html instead of spec.html, because the latter looks like a path with
3792 a MIME-type, and this confuses some software.
3794 PH/70 Catch two compiler warnings in sieve.c.
3796 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3797 function verify_get_ident() calls ip_connect() to connect a socket, but
3798 if the "connect()" function timed out, ip_connect() used to close the
3799 socket. However, verify_get_ident() also closes the socket later, and in
3800 between Exim writes to the log, which may get opened at this point. When
3801 the socket was closed in ip_connect(), the log could get the same file
3802 descriptor number as the socket. This naturally causes chaos. The fix is
3803 not to close the socket in ip_connect(); the socket should be closed by
3804 the function that creates it. There was only one place in the code where
3805 this was missing, in the iplookup router, which I don't think anybody now
3806 uses, but I've fixed it anyway.
3808 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3809 well as to direct DNS lookups. Otherwise the handling of names in host
3810 lists is inconsistent and therefore confusing.
3816 PH/01 Two changes to the default runtime configuration:
3818 (1) Move the checks for relay_from_hosts and authenticated clients from
3819 after to before the (commented out) DNS black list checks.
3821 (2) Add control=submission to the relay_from_hosts and authenticated
3822 clients checks, on the grounds that messages accepted by these
3823 statements are most likely to be submissions.
3825 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3827 (1) Generate an error if the third argument for the ${prvs expansion is
3830 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3833 (3) Reset the variables that are obtained from the first argument of
3834 ${prvscheck and used in the second argument before leaving the code,
3835 because their memory is reclaimed, so using them afterwards may do
3838 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3839 one (it's much easier than Tom thought :-).
3841 (5) Because of (4), we can now allow for the use of $prvscheck_result
3842 inside the third argument.
3844 PH/03 For some reason, the default setting of PATH when running a command from
3845 a pipe transport was just "/usr/bin". I have changed it to
3848 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3849 anything to be listed in the output from -bV.
3851 PH/05 When a filter generated an autoreply, the entire To: header line was
3852 quoted in the delivery log line, like this:
3854 => >A.N.Other <ano@some.domain> <original@ddress> ...
3856 This has been changed so that it extracts the operative address. There
3857 may be more than one such address. If so, they are comma-separated, like
3860 => >ano@some.domain,ona@other.domain <original@ddress> ...
3862 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3863 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3864 not being looked up in the rDNS to get a host name, Exim was showing the
3865 IP address twice in Received: lines, even though the IP addresses were
3866 identical. For example:
3868 Received: from [1.2.3.4] (helo=[1.2.3.4])
3870 However, if the real host name was known, it was omitting the HELO data
3871 if it matched the actual IP address. This has been tidied up so that it
3872 doesn't show the same IP address twice.
3874 PH/07 When both +timestamp and +memory debugging was on, the value given by
3875 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3876 called by the debug printing, thereby overwriting the timestamp buffer.
3877 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3879 PH/08 When the original message was included in an autoreply transport, it
3880 always said "this is a copy of the message, including all the headers",
3881 even if body_only or headers_only was set. It now gives an appropriate
3884 PH/09 Applied a patch from the Sieve maintainer which:
3886 o fixes some comments
3887 o adds the (disabled) notify extension core
3888 o adds some debug output for the result of if/elsif tests
3889 o points to the current vacation draft in the documentation
3890 and documents the missing references header update
3894 o fixes a bug in processing the envelope test (when testing
3895 multiple envelope elements, the last element determined the
3898 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3899 Electronic Mail") by including:
3901 Auto-submitted: auto-generated
3903 in the messages that it generates (bounce messages and others, such as
3904 warnings). In the case of bounce messages for non-SMTP messages, there was
3905 also a typo: it was using "Auto_submitted" (underscore instead of
3906 hyphen). Since every message generated by Exim is necessarily in response
3907 to another message, thes have all been changed to:
3909 Auto-Submitted: auto-replied
3911 in accordance with these statements in the RFC:
3913 The auto-replied keyword:
3915 - SHOULD be used on messages sent in direct response to another
3916 message by an automatic process,
3918 - MUST NOT be used on manually-generated messages,
3920 - MAY be used on Delivery Status Notifications (DSNs) and Message
3921 Disposition Notifications (MDNs),
3923 - MUST NOT be used on messages generated by automatic or periodic
3924 processes, except for messages which are automatic responses to
3927 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3928 to the default Received: header definition.
3930 PH/12 Added log selector acl_warn_skipped (default on).
3932 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3933 variables because (a) they are in the wrong storage pool and (b) even if
3934 they were copied, it wouldn't work properly because of the caching.
3936 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3937 checking when decoding. Apparently there are clients that generate
3938 overlong encoded strings. Why am I not surprised?
3940 PH/15 If the first argument of "${if match_address" was not empty, but did not
3941 contain an "@" character, Exim crashed. Now it writes a panic log message
3942 and treats the condition as false.
3944 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3946 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3947 extension "envelope-auth". The code is finished and in agreement with
3948 other implementations, but there is no documentation so far and in fact,
3949 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3950 not changing the active code.
3952 Print executed "if" and "elsif" statements when debugging is used. This
3953 helps a great deal to understand what a filter does.
3955 Document more things not specified clearly in RFC3028. I had all this
3956 sorted out, when out of a sudden new issues came to my mind. Oops."
3958 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3961 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3962 canonical form (as documented). However, after a host name lookup from
3963 the IP address, check_host() was doing a simple string comparison with
3964 addresses acquired from the DNS when checking that the found name did
3965 have the original IP as one of its addresses. Since any found IPv6
3966 addresses are likely to be in abbreviated form, the comparison could
3967 fail. Luckily, there already exists a function for doing the comparison
3968 by converting both addresses to binary, so now that is used instead of
3969 the text comparison.
3971 PH/20 There was another similar case to PH/19, when a complete host name was
3972 given in a host list; looking up its IP address could give an abbreviated
3973 form, whereas the current host's name might or might not be abbreviated.
3974 The same fix has been applied.
3980 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3981 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3984 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3985 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3987 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3989 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3990 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3991 Also, for "Auto-Submitted", treat anything other than "no" as
3992 non-personal, in accordance with RFC 3834. (Previously it treated
3993 anything starting "auto-" as non-personal.)
3995 TF/01 The control=submission/name=... option had a problem with syntax
3996 errors if the name included a slash character. The /name= option
3997 now slurps the rest of the string, so it can include any characters
3998 but it must come last in the list of options (after /sender_retain
4001 PH/05 Some modifications to the interface to the fake nameserver for the new
4009 TK/01 Added the "success_on_redirect" address verification option. See
4010 NewStuff for rationale and an example.
4012 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4014 PH/02 Patch to exigrep to allow it to work on syslog lines.
4016 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4017 fread() to skip over the body file's header line, because in Cygwin the
4018 header line is locked and is inaccessible.
4020 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4021 co-exist for some time) to make it clear that it is the Exim ID that is
4022 referenced, not the Message-ID: header line.
4024 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4025 string_format() function, because snprintf() does not exist on all
4028 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4029 ${stat: expansion item.
4031 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4032 protocol synchronization error", to keep the pedants happy.
4034 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4035 well as for IRIX systems, when gcc is being used. See the host.c source
4038 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4040 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4043 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4044 transport and to the smtp transport in LMTP mode.
4046 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4048 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4049 run a filter in a subprocess. This could lead to confusion in subsequent
4050 lookups in the parent process. There should also be a search_tidyup() at
4051 the end of the subprocess.
4053 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4054 only if the host matched helo_try_verify_hosts, which caused the
4055 verification to occur when the EHLO/HELO command was issued. The ACL just
4056 tested the remembered result. Now, if a previous verification attempt has
4057 not happened, "verify = helo" does it there and then.
4059 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4061 TK/03 Fix log output including CR from clamd.
4063 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4064 exist provoked a memory error which could cause a segfault.
4066 PH/15 Installed PCRE 6.2
4068 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4070 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4071 of the problem. Specifically, suggested +O2 rather than +O1 for the
4074 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4076 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4077 record, the port was not correctly logged when the outgoing_port log
4078 selector was set (it logged the transort's default port).
4080 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4081 fallback_hosts, and "hosts" in the smtp transport.
4083 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4084 host errors such as "Connection refused".
4086 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4087 authentication with radiusclient 0.4.9:
4089 - Error returned from rc_read_config was caught wrongly
4090 - Username/password not passed on to radius server due to wrong length.
4092 The presumption is that some radiusclient API changes for 4.51/PH/17
4093 were not taken care of correctly. The code is still untested by me (my
4094 Linux distribution still has 0.3.2 of radiusclient), but it was
4095 contributed by a Radius user.
4097 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4098 expanding the "port" option of the smtp transport.
4100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4101 while reading a MIME header. Thanks to Tom Hughes for a patch.
4103 PH/24 Include config.h inside local_scan.h so that configuration settings are
4106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4107 that in an ACL for RCPT (for example), you can examine exactly what was
4110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4111 commands, but it was not correctly comparing the address with the actual
4112 client host address. Thus, it would show the EHLO address in Received:
4113 header lines when this was not necessary.
4115 PH/27 Added the % operator to ${eval:}.
4117 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4118 it should be ignoring failures (because with -C, for example, it has lost
4119 privilege). It wasn't ignoring creation failures other than "already
4122 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4125 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4126 because an input line was too long, either on its own, or by virtue of
4127 too many continuations, the temporary file was not being removed, and the
4128 return code was incorrect.
4130 PH/31 Missing "BOOL" in function definition in filtertest.c.
4132 PH/32 Applied Sieve patches from the maintainer.
4134 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4136 PH/33 Added "verify = not_blind".
4138 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4139 Local/Makefile (with some defaults set). These are used in built scripts
4140 such as exicyclog, but they have never been used in the exim_install
4141 script (though there are many overriding facilities there). I have
4142 arranged that the exim_install script now takes note of these two
4145 PH/35 Installed configuration files for Dragonfly.
4147 PH/36 When a locally submitted message by a trusted user did not contain a
4148 From: header, and the sender address was obtained from -f or from an SMTP
4149 MAIL command, and the trusted user did not use -F to supply a sender
4150 name, $originator_name was incorrectly used when constructing a From:
4151 header. Furthermore, $originator_name was used for submission mode
4152 messages from external hosts without From: headers in a similar way,
4153 which is clearly wrong.
4155 PH/37 Added control=suppress_local_fixups.
4157 PH/38 When log_selector = +received_sender was set, and the addition of the
4158 sender made the log line's construction buffer exactly full, or one byte
4159 less than full, an overflow happened when the terminating "\n" was
4162 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4163 when the result of a list match is failure because a DNS lookup failed.
4165 PH/40 RM_COMMAND is now used in the building process.
4167 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4168 the "build-* directories that it finds.
4170 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4171 address in a domain literal was a prefix of an interface address.
4173 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4174 when verifying a sender address, unless rewrite_headers is false.
4176 PH/44 Wrote a long comment about why errors_to addresses are verified as
4177 recipients, not senders.
4179 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4180 the ratelimit ACL was added.
4182 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4184 PH/46 Added extra information about PostgreSQL errors to the error string.
4186 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4187 avoiding the need to install special zones in a real server. This is
4188 backwards compatible; if it can't find the fake resolver, it drops back.
4189 Thus, both old and new test suites can be run.
4191 TF/02 Added util/ratelimit.pl
4193 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4194 clock is set back in time.
4196 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4197 Candler <B.Candler@pobox.com>.
4199 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4200 OK for addresses that are the result of redirections.
4202 PH/48 A number of further additions for the benefit of the new test suite,
4203 including a fake gethostbyname() that interfaces to the fake DNS resolver
4206 TF/06 The fix for widen_domains has also been applied to qualify_single and
4207 search_parents which are the other dnslookup options that can cause
4210 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4213 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4215 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4216 service" if it encountered an MX record with an empty target hostname.
4217 The message is now "an MX or SRV record indicated no SMTP service".
4219 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4220 if the DNS of the sending site is misconfigured. This is quite a
4221 common situation. This change restores the behaviour of treating a
4222 helo verification defer as a failure.
4224 PH/51 If self=fail was set on a router, the bounce message did not include the
4225 actual error message.
4231 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4233 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4234 command itself ended in error, the underlying message about the transport
4235 filter timeout was being overwritten with the pipe command error. Now the
4236 underlying error message should be appended to the second error message.
4238 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4240 PH/02 Reduce the amount of output that "make" produces by default. Full output
4241 can still be requested.
4243 PH/03 The warning log line about a condition test deferring for a "warn" verb
4244 was being output only once per connection, rather than after each
4245 occurrence (because it was using the same function as for successful
4246 "warn" verbs). This seems wrong, so I have changed it.
4248 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4249 it should not have, which might have caused a crash in the right
4250 circumstances, but probably never did.
4252 PH/04 Installed a modified version of Tony Finch's patch to make submission
4253 mode fix the return path as well as the Sender: header line, and to
4254 add a /name= option so that you can make the user's friendly name appear
4257 TF/03 Added the control = fakedefer ACL modifier.
4259 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4260 Mark Lowes for thorough testing.
4262 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4264 TK/03 Merged latest SRS patch from Miles Wilton.
4266 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4267 with the definition in sysexits.h (which is #included earlier).
4268 Fortunately, Exim does not actually use EX_OK. The code used to try to
4269 preserve the sysexits.h value, by assuming that macro definitions were
4270 scanned for macro replacements. I have been disabused of this notion,
4271 so now the code just undefines EX_OK before #including unistd.h.
4273 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4274 in the smtp transport. When a block could not be written in a single
4275 write() function, the timeout was being re-applied to each part-write.
4276 This seems wrong - if the receiver was accepting one byte at a time it
4277 would take for ever. The timeout is now adjusted when this happens. It
4278 doesn't have to be particularly precise.
4280 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4281 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4283 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4284 without a domain if the check to <postmaster@domain> fails.
4286 SC/01 Eximstats: added -xls and the ability to specify output files
4287 (patch written by Frank Heydlauf).
4289 SC/02 Eximstats: use FileHandles for outputting results.
4291 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4293 SC/04 Eximstats: fixed display of large numbers with -nvr option
4295 SC/05 Eximstats: fixed merging of reports with empty tables.
4297 SC/06 Eximstats: added the -include_original_destination flag
4299 SC/07 Eximstats: removed tabs and trailing whitespace.
4301 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4303 TK/06 MBOX spool code: Add real "From " MBOX separator line
4304 so the .eml file is really in mbox format (even though
4305 most programs do not really care). Patch from Alex Miller.
4307 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4308 The latter is generated from $received_to and is only set if the
4309 message has one envelope recipient. SA can use these headers,
4310 obviously out-of-the-box. Patch from Alex Miller.
4312 PH/08 The ${def test on a variable was returning false if the variable's
4313 value was "0", contrary to what the specification has always said!
4314 The result should be true unless the variable is empty.
4316 PH/09 The syntax error of a character other than { following "${if
4317 def:variable_name" (after optional whitespace) was not being diagnosed.
4318 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4319 accidental colon was present, for example, could give incorrect results.
4321 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4322 result is used (not including appendfile, where other changes are about
4325 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4326 This involved changing a lot of size variables from int to off_t. It
4327 should work with maildirs and everything.
4329 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4330 spamd dying while we are connected to it.
4332 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4335 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4336 function for BDB 4.3.
4338 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4340 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4341 the sources for PCRE have been re-arranged and more files are now
4344 PH/15 The code I had for printing potentially long long variables in PH/11
4345 above was not the best (it lost precision). The length of off_t variables
4346 is now inspected at build time, and an appropriate printing format (%ld
4347 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4348 to be "long long int" or "long int". This is needed for the internal
4349 formatting function string_vformat().
4351 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4352 the configuration file to be ":syslog", then the script "guesses" where
4353 the logs files are, rather than using the compiled in default. In our
4354 case the guess is not the same as the compiled default, so the script
4355 suddenly stopped working when I started to use syslog. The patch checks
4356 to see if log_file_path is "". If so, it attempts to read it from exim
4357 with no configuration file to get the compiled in version, before it
4358 falls back to the previous guessing code."
4360 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4361 implementing BATV in an Exim configuration. See NewStuff for the gory
4364 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4365 Makefile that are specific to HP-UX.
4367 PH/18 If the "use_postmaster" option was set for a recipient callout together
4368 with the "random" option, the postmaster address was used as the MAIL
4369 FROM address for the random test, but not for the subsequent recipient
4370 test. It is now used for both.
4372 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4373 patch removes a few documentation additions to RFC 3028, because the
4374 latest draft now contains them. It adds the new en;ascii-case comparator
4375 and a new error check for 8bit text in MIME parts. Comparator and
4376 require names are now matched exactly. I enabled the subaddress
4377 extension, but it is not well tested yet (read: it works for me)."
4379 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4380 rework some of the code of TK/09 above to avoid the hardwired use of
4381 "%lld" and "long long". Replaced the call to snprintf() with a call to
4384 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4385 records point to non-existent hosts", "retry timeout exceeded", and
4386 "retry time not reached for any host after a long failure period".
4388 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4389 experimental DomainKeys support:
4391 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4392 (2) On an error such as an illegally used "control", the wrong name for
4393 the control was given.
4395 These problems did NOT occur unless DomainKeys support was compiled.
4397 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4399 PH/24 Added ${if match_ip condition.
4401 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4402 databases so that it will be absolutely obvious if a crash occurs in the
4403 DB library. This is a regular occurrence (often caused by mis-matched
4406 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4407 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4408 picked up on a user's system that detects such things. There doesn't seem
4409 to be a gcc warning option for this - only an attribute that has to be
4410 put on the function's prototype. It seems that in Fedora Core 4 they have
4411 set this on a number of new functions. No doubt there will be more in due
4414 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4415 specify a transport. However, if an address that was verified by such a
4416 router was the subject of a callout, Exim crashed because it tried to
4417 read the rcpt_include_affixes from the non-existent transport. Now it
4418 just assumes that the setting of that option is false. This bug was
4419 introduced by 4.51/PH/31.
4421 PH/28 Changed -d+all to exclude +memory, because that information is very
4422 rarely of interest, but it makes the output a lot bigger. People tend to
4423 do -d+all out of habit.
4425 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4426 code in os-type was giving problems when libc.so lives in lib64, like on
4429 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4430 aren't the modern standard, and it seems that some systems' include files
4431 don't always have them. Exim was already checking for some of the newer
4432 ones like T_AAAA, and defining it itself. I've added checks for all the
4433 record types that Exim uses.
4435 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4436 not automatically generating a new one, as it is supposed to. This
4437 prevented TLS from working. If the file did exist, but contained invalid
4438 data, a new version was generated, as expected. It was only the case of a
4439 non-existent file that was broken.
4441 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4442 with a change in libdomainkeys > 0.64.
4444 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4445 from DNS. If the selector record carries the flag, it now has
4446 precedence over the domain-wide flag.
4448 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4450 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4451 the use of a port name that isn't defined in /etc/services) occurred, the
4452 message was deferred as in a normal delivery, and thus remained on the
4453 spool, instead of being failed because of the mua_wrapper setting. This
4454 is now fixed, and I tidied up some of the mua_wrapper messages at the
4457 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4458 the messages in a hash of arrays rather than using individual hashes.
4459 This is a bit cleaner and results in dramatic memory savings, albeit
4460 at a slight CPU cost.
4462 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4463 as requested by Marc Sherman.
4465 SC/10 Eximstats: added histograms for user specified patterns as requested
4468 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4470 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4471 fopen() in the content-scanning modules that did not already have it.
4477 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4478 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4480 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4482 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4484 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4485 to test Sieve filters that use "vacation".
4487 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4488 that changes the way the GnuTLS parameters are stored in the cache file.
4489 The new format can be generated externally. For backward compatibility,
4490 if the data in the cache doesn't make sense, Exim assumes it has read an
4491 old-format file, and it generates new data and writes a new file. This
4492 means that you can't go back to an older release without removing the
4495 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4496 work if there are any delivery delays because "one_time" forces the
4497 parent to be marked "delivered", so its unseen clone is never tried
4498 again. For this reason, Exim now forbids the simultaneous setting of
4501 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4502 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4503 entry if you want to know the details. The fix, however, neglected to
4504 consider the case when local delivery batching is involved. The test for
4505 "previously delivered" was not happening when checking to see if an
4506 address could be batched with a previous (undelivered) one; under
4507 certain circumstances this could lead to multiple deliveries to the same
4510 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4511 in its include files, and this causes problems building Exim.
4513 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4514 header_syntax) but Exim was just ignoring anything given after a slash.
4515 In particular, this caused confusion with an attempt to use "verify =
4516 reverse_host_lookup/defer_ok". An error is now given when options are
4517 supplied for verify items that do not have them. (Maybe reverse_host_
4518 lookup should have a defer_ok option, but that's a different point.)
4520 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4521 defined by RFC 821) to 2048, because there were problems with some AUTH
4522 commands, and RFC 1869 says the size should be increased for extended
4523 SMTP commands that take arguments.
4525 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4528 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4529 "unknown" error; now it says that the functionality isn't in the binary.
4531 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4532 an address' error message when a string expansion fails (syntax or
4533 whatever). Otherwise the password may appear in the log. Following change
4534 PH/42 below, there is no longer a chance of it appearing in a bounce
4537 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4539 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4540 of its addresses was ever tried. (Bugzilla bug #2.)
4542 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4543 the result incorrectly in the debug output. (It correctly added a newline
4544 to what was transported.)
4546 TF/01 Added $received_time.
4548 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4549 commented out examples of how to interface to a virus scanner and to
4550 SpamAssassin. Also added commented examples of av_scanner and
4551 spamd_address settings.
4553 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4554 and controls are allowed in which ACLs. There were a couple of minor
4555 errors. Some of the entries in the conditions table (which is a table of
4556 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4557 negation of where the condition IS allowed.
4559 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4561 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4562 header file does not have a version number, so I've had to invent a new
4563 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4564 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4565 radiusclient), but it was contributed by a Radius user.
4567 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4568 files or directories, for OpenSSL.
4570 PH/19 When an Exim process that is running as root has to create an Exim log
4571 file, it does so in a subprocess that runs as exim:exim so as to get the
4572 ownership right at creation (otherwise, other Exim processes might see
4573 the file with the wrong ownership). There was no test for failure of this
4574 fork() call, which would lead to the process getting stuck as it waited
4575 for a non-existent subprocess. Forks do occasionally fail when resources
4576 run out. I reviewed all the other calls to fork(); they all seem to check
4579 PH/20 When checking for unexpected SMTP input at connect time (before writing
4580 the banner), Exim was not dealing correctly with a non-positive return
4581 from the read() function. If the client had disconnected by this time,
4582 the result was a log entry for a synchronization error with an empty
4583 string after "input=" when read() returned zero. If read() returned -1
4584 (an event I could not check), uninitialized data bytes were printed.
4585 There were reports of junk text (parts of files, etc) appearing after
4588 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4590 PH/22 Added support for macro redefinition, and (re)definition in between
4591 driver and ACL definitions.
4593 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4594 forgetting to use the resulting value; it was using the unexpanded value.
4596 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4597 hadn't been configured. The fix is from Juergen Kreileder, who
4598 understands it better than I do:
4600 "Here's what I see happening with three configured cyrus_sasl
4601 authenticators configured (plain, login, cram-md5):
4603 On startup auth_cyrus_sasl_init() gets called for each of these.
4604 This means three calls to sasl_listmech() without a specified mech_list.
4605 => SASL tests which mechs of all available mechs actually work
4606 => three warnings about OTP not working
4607 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4609 With the patch, sasl_listmech() also gets called three times. But now
4610 SASL's mech_list option is set to the server_mech specified in the the
4611 authenticator. Or in other words, the answer from sasl_listmech()
4612 gets limited to just the mech you're testing for (which is different
4614 => the return list contains just 'plain' or 'login', 'cram-md5' or
4615 nothing depending on the value of ob->server_mech.
4617 I've just tested the patch: Authentication still works fine,
4618 unavailable mechs specified in the exim configuration are still
4619 caught, and the auth.log warnings about OTP are gone."
4621 PH/25 When debugging is enabled, the contents of the command line are added
4622 to the debugging output, even when log_selector=+arguments is not
4625 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4626 answer is "GNU", and only if the return is "GNU/something" is the answer
4629 PH/27 $acl_verify_message is now set immediately after the failure of a
4630 verification in an ACL, and so is available in subsequent modifiers. In
4631 particular, the message can be preserved by coding like this:
4633 warn !verify = sender
4634 set acl_m0 = $acl_verify_message
4636 Previously, $acl_verify_message was set only while expanding "message"
4637 and "log_message" when a very denied access.
4639 PH/28 Modified OS/os.c-Linux with
4641 -#ifndef OS_LOAD_AVERAGE
4642 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4644 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4645 nomenclature these days.)
4647 PH/29 Installed patch from the Sieve maintainer that adds the options
4648 sieve_useraddress and sieve_subaddress to the redirect router.
4650 PH/30 In these circumstances:
4651 . Two addresses routed to the same list of hosts;
4652 . First host does not offer TLS;
4653 . First host accepts first address;
4654 . First host gives temporary error to second address;
4655 . Second host offers TLS and a TLS session is established;
4656 . Second host accepts second address.
4657 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4658 and peerdn, if requested) that were in fact used only for the second
4661 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4662 attention to any local part prefix or suffix that was matched by the
4663 router that accepted the address. It now behaves in the same way as it
4664 does for delivery: the affixes are removed from the local part unless
4665 rcpt_include_affixes is set on the transport.
4667 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4668 timeout during the DATA phase of an incoming message.
4670 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4671 applied a patch sanctioned by the Sieve maintainer.
4673 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4674 the uid or gid is negative. A case of a negative gid caused this to be
4675 noticed. The fix allows for either to be negative.
4677 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4678 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4681 PH/36 The change PH/12 above was broken. Fixed it.
4683 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4684 the grounds that routing the same address twice would always produce the
4685 same answer. This might have been true once, but it is certainly no
4686 longer true now. Routing a child address may depend on the previous
4687 routing that produced that child. Some complicated redirection strategies
4688 went wrong when messages had multiple recipients, and made Exim's
4689 behaviour dependent on the order in which the addresses were given.
4691 I have moved the duplicate checking until after the routing is complete.
4692 Exim scans the addresses that are assigned to local and remote
4693 transports, and removes any duplicates. This means that more work will be
4694 done, as duplicates will always all be routed, but duplicates are
4695 presumably rare, so I don't expect this is of any significance.
4697 For deliveries to pipes, files, and autoreplies, the duplicate checking
4698 still happens during the routing process, since they are not going to be
4701 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4702 It corrects a timeout issue with spamd. This is Ian's comment: "The
4703 background is that sometimes spamd either never reads data from a
4704 connection it has accepted, or it never writes response data. The exiscan
4705 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4706 blindly assumes that writes won't block so it may never time out."
4708 PH/39 Allow G after quota size as well as K and M.
4710 PH/40 The value set for $authenticated_id in an authenticator may not contain
4711 binary zeroes or newlines because the value is written to log lines and
4712 to spool files. There was no check on this. Now the value is run through
4713 the string_printing() function so that such characters are converted to
4714 printable escape sequences.
4716 PH/41 $message_linecount is a new variable that contains the total number of
4717 lines in the message. Compare $body_linecount, which is the count for the
4720 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4721 bounce and delay warning messages, except in certain special cases, which
4724 (a) An SMTP error message from a remote host;
4725 (b) A message specified in a :fail: redirection;
4726 (c) A message specified in a "fail" command in a system filter;
4727 (d) A message specified in a FAIL return from the queryprogram router;
4728 (e) A message specified by the cannot_route_message router option.
4730 In these cases only, Exim does include the error details in bounce and
4731 warning messages. There are also a few cases where bland messages such
4732 as "unrouteable address" or "local delivery error" are given.
4734 PH/43 $value is now also set for the "else" part of a ${run expansion.
4736 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4737 being worked on, but at least Exim now implements the latest version to
4740 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4741 process to complete was treated as a delivery failure, a timeout while
4742 writing the message to the pipe was logged, but erroneously treated as a
4743 successful delivery. Such timeouts include transport filter timeouts. For
4744 consistency with the overall process timeout, these timeouts are now
4745 treated as errors, giving rise to delivery failures by default. However,
4746 there is now a new Boolean option for the pipe transport called
4747 timeout_defer, which, if set TRUE, converts the failures into defers for
4748 both kinds of timeout. A transport filter timeout is now identified in
4751 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4752 systems where "make" and "gmake" are different, calling "gmake" at top
4753 level broke things. I've arranged for the value of $(MAKE) to be passed
4754 from the Makefile to this script so that it can call the same version of
4758 A note about Exim versions 4.44 and 4.50
4759 ----------------------------------------
4761 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4762 changes of various kinds. As a consequence, a big documentation update was
4763 needed. This delayed the release for rather longer than seemed good, especially
4764 in the light of a couple of (minor) security issues. Therefore, the changes
4765 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4766 release. So 4.44 and 4.50 are in effect two different branches that both start
4769 I have left the 4.50 change log unchanged; it contains all the changes since
4770 4.43. The change log for 4.44 is below; many of its items are identical to
4771 those for 4.50. This seems to be the most sensible way to preserve the
4772 historical information.
4778 1. Minor wording change to the doc/README.SIEVE file.
4780 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4781 computation of the current number of files was incorrect.
4783 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4784 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4787 4. Give more explanation in the error message when the command for a transport
4788 filter fails to execute.
4790 5. There are several places where Exim runs a non-Exim command in a
4791 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4792 was being done only for the command run by the queryprogram router. It is
4793 now done for all such subprocesses. The other cases are: ${run, transport
4794 filters, and the commands run by the lmtp and pipe transports.
4796 6. Added CONFIGURE_GROUP build-time option.
4798 7. Some older OS have a limit of 256 on the maximum number of file
4799 descriptors. Exim was using setrlimit() to set 1000 as a large value
4800 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4801 systems. I've change it so that if it can't get 1000, it tries for 256.
4803 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4804 was an oversight, and furthermore, ever since the addition of extra
4805 controls (e.g. 4.43/32), the checks on when to allow different forms of
4806 "control" were broken. There should now be diagnostics for all cases when a
4807 control that does not make sense is encountered.
4809 9. Added the /retain_sender option to "control=submission".
4811 10. $recipients is now available in the predata ACL (oversight).
4813 11. Tidy the search cache before the fork to do a delivery from a message
4814 received from the command line. Otherwise the child will trigger a lookup
4815 failure and thereby defer the delivery if it tries to use (for example) a
4816 cached ldap connection that the parent has called unbind on.
4818 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4819 of $address_data from the recipient verification was clobbered by the
4820 sender verification.
4822 13. The value of address_data from a sender verification is now available in
4823 $sender_address_data in subsequent conditions in the ACL statement.
4825 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4827 15. Added a new option "connect=<time>" to callout options, to set a different
4830 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4831 was its contents. (It was OK if the option was not defined at all.)
4833 17. A "Completed" log line is now written for messages that are removed from
4834 the spool by the -Mrm option.
4836 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4837 information about exactly what failed.
4839 19. Added -dd to debug only the daemon process.
4841 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4842 handles timeouts, both on the server side and network timeouts. Renamed the
4843 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4845 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4846 It is now set to "smtps".
4848 22. $host_address is now set to the target address during the checking of
4849 ignore_target_hosts.
4851 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4852 being passed; this would have caused $sender_host_name to have been used if
4853 matching the list had actually called for a host name (not very likely,
4854 since this list is usually IP addresses). A host name is now passed as
4857 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4858 code that specifies a non-restarting handler (typically sigaction() in
4859 modern systems) in an attempt to fix a rare and obscure crash bug.
4861 25. Narrowed the window for a race in the daemon that could cause it to ignore
4862 SIGCHLD signals. This is not a major problem, because they are used only to
4863 wake it up if nothing else does.
4865 26. A malformed maildirsize file could cause Exim to calculate negative values
4866 for the mailbox size or file count. Odd effects could occur as a result.
4867 The maildirsize information is now recalculated if the size or filecount
4870 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4871 support for a long time. Removed HAVE_SYS_VFS_H.
4873 28. Installed the latest version of exipick from John Jetmore.
4875 29. In an address list, if the pattern was not a regular expression, an empty
4876 subject address (from a bounce message) matched only if the pattern was an
4877 empty string. Non-empty patterns were not even tested. This was the wrong
4878 because it is perfectly reasonable to use an empty address as part of a
4879 database query. An empty address is now tested by patterns that are
4880 lookups. However, all the other forms of pattern expect the subject to
4881 contain a local part and a domain, and therefore, for them, an empty
4882 address still always fails if the pattern is not itself empty.
4884 30. Exim went into a mad DNS loop when attempting to do a callout where the
4885 host was specified on an smtp transport, and looking it up yielded more
4886 than one IP address.
4888 31. Re-factored the code for checking spool and log partition space into a
4889 function that finds that data and another that does the check. The former
4890 is then used to implement four new variables: $spool_space, $log_space,
4891 $spool_inodes, and $log_inodes.
4893 32. The RFC2047 encoding function was originally intended for short strings
4894 such as real names; it was not keeping to the 75-character limit for
4895 encoded words that the RFC imposes. It now respects the limit, and
4896 generates multiple encoded words if necessary. To be on the safe side, I
4897 have increased the buffer size for the ${rfc2047: expansion operator from
4900 33. It is now permitted to omit both strings after an "if" condition; if the
4901 condition is true, the result is "true". As before, when the second string
4902 is omitted, a false condition yields an empty string. This makes it less
4903 cumbersome to write custom ACL and router conditions.
4905 34. Failure to deliver a bounce message always caused it to be frozen, even if
4906 there was an errors_to setting on the router. The errors_to setting is now
4909 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4910 canonical form (fully expanded) before being placed in
4911 $sender_host_address.
4913 36. The table in the code that translates DNS record types into text (T_A to
4914 "A" for instance) was missing entries for NS and CNAME. It is just possible
4915 that this could have caused confusion if both these types were looked up
4916 for the same domain, because the text type is used as part of Exim's
4917 per-process caching. But the chance of anyone hitting this buglet seems
4920 37. The dnsdb lookup has been extended in a number of ways.
4922 (1) There is a new type, "zns", which walks up the domain tree until it
4923 finds some nameserver records. It should be used with care.
4925 (2) There is a new type, "mxh", which is like "mx" except that it returns
4926 just the host names, not the priorities.
4928 (3) It is now possible to give a list of domains (or IP addresses) to be
4929 looked up. The behaviour when one of the lookups defers can be
4930 controlled by a keyword.
4932 (4) It is now possible to specify the separator character for use when
4933 multiple records are returned.
4935 38. The dnslists ACL condition has been extended: it is now possible to supply
4936 a list of IP addresses and/or domains to be looked up in a particular DNS
4939 39. Added log_selector=+queue_time_overall.
4941 40. When running the queue in the test harness, wait just a tad after forking a
4942 delivery process, to get repeatability of debugging output.
4944 41. Include certificate and key file names in error message when GnuTLS fails
4945 to set them up, because the GnuTLS error message doesn't include the name
4946 of the failing file when there is a problem reading it.
4948 42. Allow both -bf and -bF in the same test run.
4950 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4952 44. The "Exiscan patch" is now merged into the mainline Exim source.
4954 45. Sometimes the final signoff response after QUIT could fail to get
4955 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4956 < 0 before doing a fflush(). This bug looks as though it goes back to the
4957 introduction of TLS in release 3.20, but "sometimes" must have been rare
4958 because the tests only now provoked it.
4960 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4961 (this can affect the format of dates).
4963 47. exim_tidydb, when checking for the continued existence of a message for
4964 which it has found a message-specific retry record, was not finding
4965 messages that were in split spool directories. Consequently, it was
4966 deleting retry records that should have stayed in existence.
4968 48. Steve fixed some bugs in eximstats.
4970 49. The SPA authentication driver was not abandoning authentication and moving
4971 on to the next authenticator when an expansion was forced to fail,
4972 contradicting the general specification for all authenticators. Instead it
4973 was generating a temporary error. It now behaves as specified.
4975 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4976 (the order specifies the preference for clients). The order is now AES256,
4977 AES128, 3DES, ARCFOUR128.
4979 51. Small patch to Sieve code - explicitly set From: when generating an
4982 52. Exim crashed if a remote delivery caused a very long error message to be
4983 recorded - for instance if somebody sent an entire SpamAssassin report back
4984 as a large number of 550 error lines. This bug was coincidentally fixed by
4985 increasing the size of one of Exim's internal buffers (big_buffer) that
4986 happened as part of the Exiscan merge. However, to be on the safe side, I
4987 have made the code more robust (and fixed the comments that describe what
4990 53. Now that there can be additional text after "Completed" in log lines (if
4991 the queue_time_overall log selector is set), a one-byte patch to exigrep
4992 was needed to allow it to recognize "Completed" as not the last thing in
4995 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4996 patch that reportedly fixes this has been added. I am not expert enough to
4997 create a test for it. This is what the patch creator wrote:
4999 "I found a little strange behaviour of ldap code when working with
5000 Windows 2003 AD Domain, where users was placed in more than one
5001 Organization Units. When I tried to give exim partial DN, the exit code
5002 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5003 But simultaneously result of request was absolutely normal ldap result,
5004 so I produce this patch..."
5006 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5007 REFERENCE, so I have modified the code to exclude the patch when that macro
5010 55. Some experimental protocols are using DNS PTR records for new purposes. The
5011 keys for these records are domain names, not reversed IP addresses. The
5012 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5013 leaves it alone. Component reversal etc. now happens only for IP addresses.
5016 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5018 57. Double the size of the debug message buffer (to 2048) so that more of very
5019 long debug lines gets shown.
5021 58. The exicyclog utility now does better if the number of log files to keep
5022 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5024 59. Two changes related to the smtp_active_hostname option:
5026 (1) $smtp_active_hostname is now available as a variable.
5027 (2) The default for smtp_banner uses $smtp_active_hostname instead
5028 of $primary_hostname.
5030 60. The host_aton() function is supposed to be passed a string that is known
5031 to be a valid IP address. However, in the case of IPv6 addresses, it was
5032 not checking this. This is a hostage to fortune. Exim now panics and dies
5033 if the condition is not met. A case was found where this could be provoked
5034 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5035 components; fortuitously, this particular loophole had already been fixed
5036 by change 4.50/55 above.
5038 If there are any other similar loopholes, the new check in host_aton()
5039 itself should stop them being exploited. The report I received stated that
5040 data on the command line could provoke the exploit when Exim was running as
5041 exim, but did not say which command line option was involved. All I could
5042 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5043 running as the user.
5046 61. There was a buffer overflow vulnerability in the SPA authentication code
5047 (which came originally from the Samba project). I have added a test to the
5048 spa_base64_to_bits() function which I hope fixes it.
5051 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5052 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5054 63. The daemon start-up calls getloadavg() while still root for those OS that
5055 need the first call to be done as root, but it missed one case: when
5056 deliver_queue_load_max is set with deliver_drop_privilege. This is
5057 necessary for the benefit of the queue runner, because there is no re-exec
5058 when deliver_drop_privilege is set.
5060 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5061 This has been fixed.
5063 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5064 was in use, was not putting the data itself into the right store pool;
5065 consequently, it could be overwritten for a subsequent message in the same
5066 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5069 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5071 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5072 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5073 the function were treating the return as a boolean value, which happened to
5074 work because 0=false and not-0=true, but is not correct code.
5076 68. The host_aton() function was not handling scoped IPv6 addresses (those
5077 with, for example, "%eth0" on the end) correctly.
5079 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5080 negated items (that is, ~something) in unsigned ints. Some compilers
5081 apparently mutter when there is no cast.
5083 70. If an address verification called from an ACL failed, and did not produce a
5084 user-specific message (i.e. there was only a "system" message), nothing was
5085 put in $acl_verify_message. In this situation, it now puts the system
5088 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5089 session; change 4.31/43 added the unwanted input to the log line - except
5090 that it did not do this in the start of session case. It now does.
5092 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5093 This is wrong and can cause the other end to generate a synchronization
5094 error if it is another Exim or anything else that does the synchronization
5095 check. A QUIT command is no longer sent after a timeout.
5097 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5098 during host lookups.
5100 74. The defer_ok option of callout verification was not working if it was used
5101 when verifying addresses in header lines, that is, for this case:
5103 verify = header_sender/callout=defer_ok
5105 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5106 those file descriptors could be used for SMTP connections. If anything
5107 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5108 could be sent to the SMTP client, causing chaos. The daemon now opens
5109 stdin, stdout, and stderr to /dev/null when it puts itself into the
5112 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5113 log by default. The user can override this with suitable Perl magic.
5115 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5116 the log message when discard triggers, was not working for the DATA ACL or
5117 for the non-SMTP ACL.
5119 78. Error message wording change in sieve.c.
5121 79. If smtp_accept_max_per_host was set, the number of connections could be
5122 restricted to fewer than expected, because the daemon was trying to set up
5123 a new connection before checking whether the processes handling previous
5124 connections had finished. The check for completed processes is now done
5125 earlier. On busy systems, this bug wouldn't be noticed because something
5126 else would have woken the daemon, and it would have reaped the completed
5129 80. If a message was submitted locally by a user whose login name contained one
5130 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5131 It caused a spool format error. I have fixed the spool reading code. A
5132 related problem was that the "from" clause in the Received: line became
5133 illegal because of the space(s). It is now covered by ${quote_local_part.
5135 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5138 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5140 83. Patch from Sieve maintainer to fix unterminated string problem in
5141 "vacation" handling.
5143 84. Some minor changes to the Linux configuration files to help with other
5144 OS variants using glibc.
5146 85. One more patch for Sieve to update vacation handling to latest spec.
5149 ----------------------------------------------------
5150 See the note above about the 4.44 and 4.50 releases.
5151 ----------------------------------------------------
5157 1. Change 4.43/35 introduced a bug that caused file counts to be
5158 incorrectly computed when quota_filecount was set in an appendfile
5161 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5162 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5165 3. Give more explanation in the error message when the command for a transport
5166 filter fails to execute.
5168 4. There are several places where Exim runs a non-Exim command in a
5169 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5170 was being done only for the command run by the queryprogram router. It is
5171 now done for all such subprocesses. The other cases are: ${run, transport
5172 filters, and the commands run by the lmtp and pipe transports.
5174 5. Some older OS have a limit of 256 on the maximum number of file
5175 descriptors. Exim was using setrlimit() to set 1000 as a large value
5176 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5177 systems. I've change it so that if it can't get 1000, it tries for 256.
5179 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5180 was an oversight, and furthermore, ever since the addition of extra
5181 controls (e.g. 4.43/32), the checks on when to allow different forms of
5182 "control" were broken. There should now be diagnostics for all cases when a
5183 control that does not make sense is encountered.
5185 7. $recipients is now available in the predata ACL (oversight).
5187 8. Tidy the search cache before the fork to do a delivery from a message
5188 received from the command line. Otherwise the child will trigger a lookup
5189 failure and thereby defer the delivery if it tries to use (for example) a
5190 cached ldap connection that the parent has called unbind on.
5192 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5193 of $address_data from the recipient verification was clobbered by the
5194 sender verification.
5196 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5197 was its contents. (It was OK if the option was not defined at all.)
5199 11. A "Completed" log line is now written for messages that are removed from
5200 the spool by the -Mrm option.
5202 12. $host_address is now set to the target address during the checking of
5203 ignore_target_hosts.
5205 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5206 being passed; this would have caused $sender_host_name to have been used if
5207 matching the list had actually called for a host name (not very likely,
5208 since this list is usually IP addresses). A host name is now passed as
5211 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5212 code that specifies a non-restarting handler (typically sigaction() in
5213 modern systems) in an attempt to fix a rare and obscure crash bug.
5215 15. Narrowed the window for a race in the daemon that could cause it to ignore
5216 SIGCHLD signals. This is not a major problem, because they are used only to
5217 wake it up if nothing else does.
5219 16. A malformed maildirsize file could cause Exim to calculate negative values
5220 for the mailbox size or file count. Odd effects could occur as a result.
5221 The maildirsize information is now recalculated if the size or filecount
5224 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5225 support for a long time. Removed HAVE_SYS_VFS_H.
5227 18. Updated exipick to current release from John Jetmore.
5229 19. Allow an empty sender to be matched against a lookup in an address list.
5230 Previously the only cases considered were a regular expression, or an
5233 20. Exim went into a mad DNS lookup loop when doing a callout where the
5234 host was specified on the transport, if the DNS lookup yielded more than
5237 21. The RFC2047 encoding function was originally intended for short strings
5238 such as real names; it was not keeping to the 75-character limit for
5239 encoded words that the RFC imposes. It now respects the limit, and
5240 generates multiple encoded words if necessary. To be on the safe side, I
5241 have increased the buffer size for the ${rfc2047: expansion operator from
5244 22. Failure to deliver a bounce message always caused it to be frozen, even if
5245 there was an errors_to setting on the router. The errors_to setting is now
5248 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5249 canonical form (fully expanded) before being placed in
5250 $sender_host_address.
5252 24. Updated eximstats to version 1.33
5254 25. Include certificate and key file names in error message when GnuTLS fails
5255 to set them up, because the GnuTLS error message doesn't include the name
5256 of the failing file when there is a problem reading it.
5258 26. Expand error message when OpenSSL has problems setting up cert/key files.
5261 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5262 (this can affect the format of dates).
5264 28. exim_tidydb, when checking for the continued existence of a message for
5265 which it has found a message-specific retry record, was not finding
5266 messages that were in split spool directories. Consequently, it was
5267 deleting retry records that should have stayed in existence.
5269 29. eximstats updated to version 1.35
5270 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5271 1.35 - bugfix such that pie charts by volume are generated correctly
5273 30. The SPA authentication driver was not abandoning authentication and moving
5274 on to the next authenticator when an expansion was forced to fail,
5275 contradicting the general specification for all authenticators. Instead it
5276 was generating a temporary error. It now behaves as specified.
5278 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5279 (the order specifies the preference for clients). The order is now AES256,
5280 AES128, 3DES, ARCFOUR128.
5282 31. Small patch to Sieve code - explicitly set From: when generating an
5285 32. Exim crashed if a remote delivery caused a very long error message to be
5286 recorded - for instance if somebody sent an entire SpamAssassin report back
5287 as a large number of 550 error lines. This bug was coincidentally fixed by
5288 increasing the size of one of Exim's internal buffers (big_buffer) that
5289 happened as part of the Exiscan merge. However, to be on the safe side, I
5290 have made the code more robust (and fixed the comments that describe what
5293 33. Some experimental protocols are using DNS PTR records for new purposes. The
5294 keys for these records are domain names, not reversed IP addresses. The
5295 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5296 leaves it alone. Component reversal etc. now happens only for IP addresses.
5299 34. The host_aton() function is supposed to be passed a string that is known
5300 to be a valid IP address. However, in the case of IPv6 addresses, it was
5301 not checking this. This is a hostage to fortune. Exim now panics and dies
5302 if the condition is not met. A case was found where this could be provoked
5303 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5304 components; fortuitously, this particular loophole had already been fixed
5305 by change 4.50/55 or 4.44/33 above.
5307 If there are any other similar loopholes, the new check in host_aton()
5308 itself should stop them being exploited. The report I received stated that
5309 data on the command line could provoke the exploit when Exim was running as
5310 exim, but did not say which command line option was involved. All I could
5311 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5312 running as the user.
5315 35. There was a buffer overflow vulnerability in the SPA authentication code
5316 (which came originally from the Samba project). I have added a test to the
5317 spa_base64_to_bits() function which I hope fixes it.
5320 36. The daemon start-up calls getloadavg() while still root for those OS that
5321 need the first call to be done as root, but it missed one case: when
5322 deliver_queue_load_max is set with deliver_drop_privilege. This is
5323 necessary for the benefit of the queue runner, because there is no re-exec
5324 when deliver_drop_privilege is set.
5326 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5327 was in use, was not putting the data itself into the right store pool;
5328 consequently, it could be overwritten for a subsequent message in the same
5329 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5332 38. Sometimes the final signoff response after QUIT could fail to get
5333 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5334 < 0 before doing a fflush(). This bug looks as though it goes back to the
5335 introduction of TLS in release 3.20, but "sometimes" must have been rare
5336 because the tests only now provoked it.
5342 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5343 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5344 Now it returns an integer. A number of calls were still expecting a T/F
5345 return. Fortuitously, in all cases, the tests worked in OK situations,
5346 which is the norm. However, things would have gone wrong on any write
5347 failures on the smtp file descriptor. This function is used when sending
5348 messages over SMTP and also when doing verify callouts.
5350 2. When Exim is called to do synchronous delivery of a locally submitted
5351 message (the -odf or -odi options), it no longer closes stderr before doing
5354 3. Implemented the mua_wrapper option.
5356 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5358 5. Implemented the functions header_remove(), header_testname(),
5359 header_add_at_position(), and receive_remove_recipient(), and exported them
5362 6. If an ACL "warn" statement specified the addition of headers, Exim already
5363 inserted X-ACL-Warn: at the start if there was no header name. However, it
5364 was not making this test for the second and subsequent header lines if
5365 there were newlines in the string. This meant that an invalid header could
5366 be inserted if Exim was badly configured.
5368 7. Allow an ACL "warn" statement to add header lines at the start or after all
5369 the Received: headers, as well as at the end.
5371 8. Added the rcpt_4xx retry error code.
5373 9. Added postmaster_mailfrom=xxx to callout verification option.
5375 10. Added mailfrom=xxxx to the callout verification option, for verify=
5378 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5379 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5381 12. Inserted some casts to stop certain compilers warning when using pointer
5382 differences as field lengths or precisions in printf-type calls (mostly
5383 affecting debugging statements).
5385 13. Added optional readline() support for -be (dynamically loaded).
5387 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5388 same clock tick as a message's arrival, so that its received time was the
5389 same as the "first fail" time on the retry record, and that message
5390 remained on the queue past the ultimate address timeout, every queue runner
5391 would try a delivery (because it was past the ultimate address timeout) but
5392 after another failure, the ultimate address timeout, which should have then
5393 bounced the address, did not kick in. This was a "< instead of <=" error;
5394 in most cases the first failure would have been in the next clock tick
5395 after the received time, and all would be well.
5397 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5398 being recognized when the domain list was tested by the match_domain
5399 condition in an expansion string.
5401 16. Added the ${str2b64: operator.
5403 17. Exim was always calling setrlimit() to set a large limit for the number of
5404 processes, without checking whether the existing limit was already
5405 adequate. (It did check for the limit on file descriptors.) Furthermore,
5406 errors from getrlimit() and setrlimit() were being ignored. Now they are
5407 logged to the main and panic logs, but Exim does carry on, to try to do its
5408 job under whatever limits there are.
5410 18. Imported PCRE 5.0.
5412 19. Trivial typo in log message " temporarily refused connection" (the leading
5415 20. If the log selector return_path_on_delivery was set and an address was
5416 redirected to /dev/null, the delivery process crashed because it assumed
5417 that a return path would always be set for a "successful" delivery. In this
5418 case, the whole delivery is bypassed as an optimization, and therefore no
5421 21. Internal re-arrangement: the function for sending a challenge and reading
5422 a response while authentication was assuming a zero-terminated challenge
5423 string. It's now changed to take a pointer and a length, to allow for
5424 binary data in such strings.
5426 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5428 23. Exim was not respecting finduser_retries when seeking the login of the
5429 uid under which it was called; it was always trying 10 times. (The default
5430 setting of finduser_retries is zero.) Also, it was sleeping after the final
5431 failure, which is pointless.
5433 24. Implemented tls_on_connect_ports.
5435 25. Implemented acl_smtp_predata.
5437 26. If the domain in control=submission is set empty, Exim assumes that the
5438 authenticated id is a complete email address when it generates From: or
5439 Sender: header lines.
5441 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5442 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5443 chown and chgrp in /bin and hostname in /usr/bin.
5445 28. Exim was keeping the "process log" file open after each use, just as it
5446 does for the main log. This opens the possibility of it remaining open for
5447 long periods when the USR1 signal hits a daemon. Occasional processlog
5448 errors were reported, that could have been caused by this. Anyway, it seems
5449 much more sensible not to leave this file open at all, so that is what now
5452 29. The long-running daemon process does not normally write to the log once it
5453 has entered its main loop, and it closes the log before doing so. This is
5454 so that log files can straightforwardly be renamed and moved. However,
5455 there are a couple of unusual error situations where the daemon does write
5456 log entries, and I had neglected to close the log afterwards.
5458 30. The text of an SMTP error response that was received during a remote
5459 delivery was being truncated at 512 bytes. This is too short for some of
5460 the long messages that one sometimes sees. I've increased the limit to
5463 31. It is now possible to make retry rules that apply only when a message has a
5464 specific sender, in particular, an empty sender.
5466 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5467 it possible to be selective about when SMTP synchronization is enforced.
5469 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5471 32. Implemented hosts_connection_nolog.
5473 33. Added an ACL for QUIT.
5475 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5478 35. Added mailbox_size and mailbox_filecount to appendfile.
5480 36. Added control = no_multiline_responses to ACLs.
5482 37. There was a bug in the logic of the code that waits for the clock to tick
5483 in the case where the clock went backwards by a substantial amount such
5484 that the microsecond fraction of "now" was more than the microsecond
5485 fraction of "then" (but the whole seconds number was less).
5487 38. Added support for the libradius Radius client library this is found on
5488 FreeBSD (previously only the radiusclient library was supported).
5494 1. When certain lookups returned multiple values in the form name=value, the
5495 quoting of the values was not always being done properly. Specifically:
5496 (a) If the value started with a double quote, but contained no whitespace,
5498 (b) If the value contained whitespace other than a space character (i.e.
5499 tabs or newlines or carriage returns) it was not quoted.
5500 This fix has been applied to the mysql and pgsql lookups by writing a
5501 separate quoting function and calling it from the lookup code. The fix
5502 should probably also be applied to nisplus, ibase and oracle lookups, but
5503 since I cannot test any of those, I have not disturbed their existing code.
5505 2. A hit in the callout cache for a specific address caused a log line with no
5506 reason for rejecting RCPT. Now it says "Previous (cached) callout
5507 verification failure".
5509 3. There was an off-by-one bug in the queryprogram router. An over-long
5510 return line was truncated at 256 instead of 255 characters, thereby
5511 overflowing its buffer with the terminating zero. As well as fixing this, I
5512 have increased the buffer size to 1024 (and made a note to document this).
5514 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5515 when Exim is waiting for an SMTP response from a remote server, Exim
5516 restarts its select() call on the socket, thereby resetting its timeout.
5517 This is not a problem when such interrupts are rare. Somebody set up a cron
5518 job to run exiwhat every 2 minutes, which is less than the normal select()
5519 timeout (5 or 10 minutes). This meant that the select() timeout never
5520 kicked in because it was always reset. I have fixed this by comparing the
5521 time when an interrupt arrives with the time at the start of the first call
5522 to select(). If more time than the timeout has elapsed, the interrupt is
5523 treated as a timeout.
5525 5. Some internal re-factoring in preparation for the addition of Sieve
5526 extensions (by MH). In particular, the "personal" test is moved to a
5527 separate function, and given an option for scanning Cc: and Bcc: (which is
5528 not set for Exim filters).
5530 6. When Exim created an email address using the login of the caller as the
5531 local part (e.g. when creating a From: or Sender: header line), it was not
5532 quoting the local part when it contained special characters such as @.
5534 7. Installed new OpenBSD configuration files.
5536 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5537 try to make them clearer.
5539 9. Callout options, other than the timeout value, were being ignored when
5540 verifying sender addresses in header lines. For example, when using
5542 verify = header_sender/callout=no_cache
5544 the cache was (incorrectly) being used.
5546 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5547 only the Cygwin environment.
5549 11. When return_path_on_delivery was set as a log selector, if different remote
5550 addresses in the same message used different return paths and parallel
5551 remote delivery occurred, the wrong values would sometimes be logged.
5552 (Whenever a remote delivery process finished, the return path value from
5553 the most recently started remote delivery process was logged.)
5555 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5556 lines when AUTH and/or TLS are in use. This is the "received protocol"
5557 field. Exim used to use "asmtp" for authenticated SMTP, without any
5558 indication (in the protocol name) for TLS use. Now it follows the RFC and
5559 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5560 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5561 names appear in log lines as well as in Received: header lines.
5563 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5564 extensions, and comparison tests, and to fix some bugs.
5566 14. Changes to the "personal" filter test:
5568 (1) The test was buggy in that it was just doing the equivalent of
5569 "contains" tests on header lines. For example, if a user's address was
5570 anne@some.where, the "personal" test would incorrectly be true for
5572 To: susanne@some.where
5574 This test is now done by extracting each address from the header in turn,
5575 and checking the entire address. Other tests that are part of "personal"
5576 are now done using regular expressions (for example, to check local parts
5577 of addresses in From: header lines).
5579 (2) The list of non-personal local parts in From: addresses has been
5580 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5581 taken from the Sieve specification recommendations.
5583 (3) If the message contains any header line starting with "List-" it is
5584 treated as non-personal.
5586 (4) The test for "circular" in the Subject: header line has been removed
5587 because it now seems ill-conceived.
5589 15. Minor typos in src/EDITME comments corrected.
5591 16. Installed latest exipick from John Jetmore.
5593 17. If headers_add on a router specified a text string that was too long for
5594 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5595 of string_sprintf() is now avoided.
5597 18. $message_body_size was not set (it was always zero) when running the DATA
5598 ACL and the local_scan() function.
5600 19. For the "mail" command in an Exim filter, no default was being set for
5601 the once_repeat time, causing a random time value to be used if "once" was
5602 specified. (If the value happened to be <= 0, no repeat happened.) The
5603 default is now 0s, meaning "never repeat". The "vacation" command was OK
5604 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5605 (I found it when inspecting the code).
5607 20. There is now an overall timeout for performing a callout verification. It
5608 defaults to 4 times the callout timeout, which applies to individual SMTP
5609 commands during the callout. The overall timeout applies when there is more
5610 than one host that can be tried. The timeout is checked before trying the
5611 next host. This prevents very long delays if there are a large number of
5612 hosts and all are timing out (e.g. when the network connections are timing
5613 out). The value of the overall timeout can be changed by specifying an
5614 additional sub-option for "callout", called "maxwait". For example:
5616 verify = sender/callout=5s,maxwait=20s
5618 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5619 to the end before writing, but this should make it even safer).
5621 22. Exim was forgetting that it had advertised PIPELINING for the second and
5622 subsequent messages on an SMTP connection. It was also not resetting its
5623 memory on STARTTLS and an internal HELO.
5625 23. When Exim logs an SMTP synchronization error within a session, it now
5626 records whether PIPELINING has been advertised or not.
5628 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5629 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5630 rather than long int.
5632 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5634 26. Added the never_mail option to autoreply.
5640 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5641 crash if the getsockname() call failed; this can happen if a connection is
5642 closed very soon after it is established. The problem was simply in the
5643 order in which certain operations were done, causing Exim to try to write
5644 to the SMTP stream before it had set up the file descriptor. The bug has
5645 been fixed by making things happen in the correct order.
5651 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5652 before the connection was closed, thus losing the rejection response.
5654 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5655 some early Solaris releases, but causes trouble in current releases where
5656 socklen_t is defined.
5658 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5661 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5664 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5665 that releasing the top of it at the end releases what was used for sub-
5666 expansions (unless the block got too big). However, discard this block if
5667 the first thing is a variable or header, so that we can use its block when
5668 it is dynamic (useful for very large $message_headers, for example).
5670 6. Lookups now cache *every* query, not just the most recent. A new, separate
5671 store pool is used for this. It can be recovered when all lookup caches are
5672 flushed. Lookups now release memory at the end of their result strings.
5673 This has involved some general refactoring of the lookup sources.
5675 7. Some code has been added to the store_xxx() functions to reduce the amount
5676 of flapping under certain conditions.
5678 8. log_incoming_interface used to affect only the <= reception log lines. Now
5679 it causes the local interface and port to be added to several more SMTP log
5680 lines, for example "SMTP connection from", and rejection lines.
5682 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5684 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5686 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5687 could be overwritten at the end of the current message (or the start of a
5688 new message if it was set in a HELO ACL). The value is now preserved for
5689 the duration of the SMTP connection.
5691 12. If a transport had a headers_rewrite setting, and a matching header line
5692 contained an unqualified address, that address was qualified, even if it
5693 did not match any rewriting rules. The underlying bug was that the values
5694 of the flags that permit the existence of unqualified sender and recipient
5695 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5696 non-local messages, and by -bnq for local messages) were not being
5697 preserved with the message after it was received.
5699 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5700 "next input=" as part of the text comprising the host identity instead of
5701 the correct text. The code was using the same buffer for two different
5702 strings. However, depending on which order the printing function evaluated
5703 its arguments, the bug did not always show up. Under Linux, for example, my
5704 test suite worked just fine.
5706 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5707 doesn't work with some older versions of Perl. It has been changed to "my",
5708 which in any case is probably the better facility to use.
5710 15. A really picky compiler found some instances of statements for creating
5711 error messages that either had too many or two few arguments for the format
5714 16. The size of the buffer for calls to the DNS resolver has been increased
5715 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5716 for addresses that have a lot of PTR records. This alleviates a problem; it
5717 does not fully solve it.
5719 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5720 buffer now truncates the list and logs the incident, which is the same
5721 action as happens when Exim is looking up a host name and its aliases.
5722 Previously in this situation something unpredictable would happen;
5723 sometimes it was "internal error: store_reset failed".
5725 18. If a server dropped the connection unexpectedly when an Exim client was
5726 using GnuTLS and trying to read a response, the client delivery process
5727 crashed while trying to generate an error log message.
5729 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5730 string, for example:
5732 warn message = H1: something\nH2: something
5734 the text was added as a single header line from Exim's point of view
5735 though it ended up OK in the delivered message. However, searching for the
5736 second and subsequent header lines using $h_h2: did not work. This has been
5737 fixed. Similarly, if a system filter added multiple headers in this way,
5738 the routers could not see them.
5740 20. Expanded the error message when iplsearch is called with an invalid key to
5741 suggest using net-iplsearch in a host list.
5743 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5744 ACLs are no longer actually imposed (and a message to that effect is
5747 22. If a "gecos" field in a passwd entry contained escaped characters, in
5748 particular, if it contained a \" sequence, Exim got it wrong when building
5749 a From: or a Sender: header from that name. A second bug also caused
5750 incorrect handling when an unquoted " was present following a character
5751 that needed quoting.
5753 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5754 was not being matched caselessly.
5756 24. Arranged for all hyphens in the exim.8 source to be escaped with
5759 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5760 properly. Recipient callout cache records were still being keyed to include
5761 the sender, even when use_sender was set false. This led to far more
5762 callouts that were necessary. The sender is no longer included in the key
5763 when use_sender is false.
5765 26. Added "control = submission" modifier to ACLs.
5767 27. Added the ${base62d: operator to decode base 62 numbers.
5769 28. dnsdb lookups can now access SRV records.
5771 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5772 the configuration file.
5774 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5775 (-v) mode. This makes the output for a verbose queue run more intelligible.
5777 31. Added a use_postmaster feature to recipient callouts.
5779 32. Added the $body_zerocount variable, containing the number of binary zero
5780 bytes in the message body.
5782 33. The time of last modification of the "new" subdirectory is now used as the
5783 "mailbox time last read" when there is a quota error for a maildir
5786 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5788 35. Added +ignore_unknown as a special item in host lists.
5790 36. Code for decoding IPv6 addresses in host lists is now included, even if
5791 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5792 address was recognized as an IP address, but was then not correctly decoded
5793 into binary, causing unexpected and incorrect effects when compared with
5800 1. Very minor rewording of debugging text in manualroute to say "list of
5801 hosts" instead of "hostlist".
5803 2. If verify=header_syntax was set, and a header line with an unqualified
5804 address (no domain) and a large number of spaces between the end of the
5805 name and the colon was received, the reception process suffered a buffer
5806 overflow, and (when I tested it) crashed. This was caused by some obsolete
5807 code that should have been removed. The fix is to remove it!
5809 3. When running in the test harness, delay a bit after writing a bounce
5810 message to get a bit more predictability in the log output.
5812 4. Added a call to search_tidyup() just before forking a reception process. In
5813 theory, someone could use a lookup in the expansion of smtp_accept_max_
5814 per_host which, without the tidyup, could leave open a database connection.
5816 5. Added the variables $recipient_data and $sender_data which get set from a
5817 lookup success in an ACL "recipients" or "senders" condition, or a router
5818 "senders" option, similar to $domain_data and $local_part_data.
5820 6. Moved the writing of debug_print from before to after the "senders" test
5823 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5824 problems for message scanning, either using a data ACL, or using
5825 local_scan() because the Received: header was not generated till after they
5826 were called (in order to set the time as the time of reception completion).
5827 I have revised the way this works. The header is now generated after the
5828 body is received, but before the ACL or local_scan() are called. After they
5829 are run, the timestamp in the header is updated.
5835 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5836 before starting a queue runner without re-exec. This happened only when
5837 deliver_drop_privilege was set or when the Exim user was set to root. The
5838 effect of the bug was that timeouts during subsequent deliveries caused
5839 crashes instead of being properly handled. The handler is now left at its
5840 default (and expected) setting.
5842 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5843 message, again when deliver_drop_privilege is set or Exim is run as root.
5844 The bug described in (1) was not present in this case, but the tidying up
5845 of the other signals was missing. I have made the two cases consistent.
5847 3. The ignore_target_hosts setting on a manualroute router was being ignored
5848 for hosts that were looked up using the /MX notation.
5850 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5853 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5854 operated on the sender address. After changing the $sender_address to <>
5855 for the sender address verify, Exim was re-instated it as the original
5856 (before rewriting) address, but remembering that it had rewritten it, so it
5857 wasn't rewriting it again. This bug also had the effect of breaking the
5858 sender address verification caching when the sender address was rewritten.
5860 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5861 This has been changed so that if the ip literal address matches
5862 ignore_target_hosts, the router declines.
5864 7. Added expansion conditions match_domain, match_address, and match_local_
5865 part (NOT match_host).
5867 8. The placeholder for the Received: header didn't have a length field set.
5869 9. Added code to Exim itself and to exim_lock to test for a specific race
5870 condition that could lead to file corruption when using MBX delivery. The
5871 issue is with the lockfile that is created in /tmp. If this file is removed
5872 after a process has opened it but before that process has acquired a lock,
5873 there is the potential for a second process to recreate the file and also
5874 acquire a lock. This could lead to two Exim processes writing to the file
5875 at the same time. The added code performs the same test as UW imapd; it
5876 checks after acquiring the lock that its file descriptor still refers to
5877 the same named file.
5879 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5880 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5883 11. Added the smtp_active_hostname option. If used, this will typically be made
5884 to depend on the incoming interface address. Because $interface_address is
5885 not set up until the daemon has forked a reception process, error responses
5886 that can happen earlier (such as "too many connections") no longer contain
5889 12. If an expansion in a condition on a "warn" statement fails because a lookup
5890 defers, the "warn" statement is abandoned, and the next ACL statement is
5891 processed. Previously this caused the whole ACL to be aborted.
5893 13. Added the iplsearch lookup type.
5895 14. Added ident_timeout as a log selector.
5897 15. Added tls_certificate_verified as a log selector.
5899 16. Added a global option tls_require_ciphers (compare the smtp transport
5900 option of the same name). This controls incoming TLS connections.
5902 17. I finally figured out how to make tls_require_ciphers do a similar thing
5903 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5904 before starting the TLS session.
5906 18. Tabs are now shown as \t in -bP output.
5908 19. If the log selector return_path_on_delivery was set, Exim crashed when
5909 bouncing a message because it had too many Received: header lines.
5911 20. If two routers both had headers_remove settings, and the first one included
5912 a superfluous trailing colon, the final name in the first list and the
5913 first name in the second list were incorrectly joined into one item (with a
5914 colon in the middle).
5920 1. Added -C and -D options to the exinext utility, mainly to make it easier
5921 to include in the automated testing, but these could be helpful when
5922 multiple configurations are in use.
5924 2. The exinext utility was not formatting the output nicely when there was
5925 an alternate port involved in the retry record key, nor when there was a
5926 message id as well (for retries that were specific to a specific message
5927 and a specific host). It was also confused by IPv6 addresses, because of
5928 the additional colons they contain. I have fixed the IPv4 problem, and
5929 patched it up to do a reasonable job for IPv6.
5931 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5932 delivery, the log line now contains "pipelined" if PIPELINING was used.
5934 4. An SMTP transport process used to panic and die if the bind() call to set
5935 an explicit outgoing interface failed. This has been changed; it is now
5936 treated in the same way as a connect() failure.
5938 5. A reference to $sender_host_name in the part of a conditional expansion
5939 that was being skipped was still causing a DNS lookup. This no longer
5942 6. The def: expansion condition was not recognizing references to header lines
5943 that used bh_ and bheader_.
5945 7. Added the _cache feature to named lists.
5947 8. The code for checking quota_filecount in the appendfile transport was
5948 allowing one more file than it should have been.
5950 9. For compatibility with Sendmail, the command line option
5958 and sets the incoming protocol and host name (for trusted callers). The
5959 host name and its colon can be omitted when only the protocol is to be set.
5960 Note the Exim already has two private options, -pd and -ps, that refer to
5961 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5962 "s", but I don't think that's a major issue.
5964 10. A number of refactoring changes to the code, none of which should affect
5967 (a) The number of logging options was getting close to filling up the
5968 32-bit word that was used as a bit map. I have split them into two classes:
5969 those that are passed in the argument to log_write(), and those that are
5970 only ever tested independently outside of that function. These are now in
5971 separate 32-bit words, so there is plenty of room for expansion again.
5972 There is no change in the user interface or the logging behaviour.
5974 (b) When building, for example, log lines, the code previously used a
5975 macro that called string_cat() twice, in order to add two strings. This is
5976 not really sufficiently general. Furthermore, there was one instance where
5977 it was actually wrong because one of the argument was used twice, and in
5978 one call a function was used. (As it happened, calling the function twice
5979 did not affect the overall behaviour.) The macro has been replaced by a
5980 function that can join an arbitrary number of extra strings onto a growing
5983 (c) The code for expansion conditions now uses a table and a binary chop
5984 instead of a serial search (which was left over from when there were very
5985 few conditions). Also, it now recognizes conditions like "pam" even when
5986 the relevant support is not compiled in: a suitably worded error message is
5987 given if an attempt is made to use such a condition.
5989 11. Added ${time_interval:xxxxx}.
5991 12. A bug was causing one of the ddress fields not to be passed back correctly
5992 from remote delivery subprocesses. The field in question was not being
5993 subsequently used, so this caused to problems in practice.
5995 13. Added new log selectors queue_time and deliver_time.
5997 14. Might have fixed a bug in maildirsizefile handling that threw up
5998 "unexpected character" debug warnings, and recalculated the data
5999 unnecessarily. In any case, I expanded the warning message to give more
6002 15. Added the message "Restricted characters in address" to the statements in
6003 the default ACL that block characters like @ and % in local parts.
6005 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6006 Three changes have been made:
6008 (a) There was a serious bug; a negative response to MAIL caused the whole
6009 recipient domain to be cached as invalid, thereby blocking all messages
6010 to all local parts at the same domain, from all senders. This bug has
6011 been fixed. The domain is no longer cached after a negative response to
6012 MAIL if the sender used is not empty.
6014 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6017 (c) A new callout option, "use_sender" has been added for people who want
6018 the modified behaviour.
6024 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6027 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6028 indeed breaks things for older releases.
6030 3. Added additional logging to the case where there is a problem reading data
6031 from a filter that is running in a subprocess using a pipe, in order to
6032 try to track down a specific problem.
6034 4. Testing facility fudge: when running in the test harness and attempting
6035 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6036 getting "No route to host". Convert this to a timeout.
6038 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6041 6. Some OS don't have socklen_t but use size_t instead. This affects the
6042 fifth argument of getsockopt() amongst other things. This is now
6043 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6044 can be set for individual OS. I have set it for SunOS5, OSF1, and
6045 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6046 some earlier ones do not.
6048 7. Change 4.30/15 was not doing the test caselessly.
6050 8. The standard form for an IPv6 address literal was being rejected by address
6051 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6052 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6053 this, as well as the form without the "IPv6" on the front (but only when
6054 address literals are enabled, of course).
6056 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6058 10. Exim crashed if a message with an empty sender address specified by -f
6059 encountered a router with an errors_to setting. This could be provoked only
6060 by a command such as
6064 where an empty string was supplied; "<>" did not hit this bug.
6066 11. Installed PCRE release 4.5.
6068 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6069 remained set. It is now erased.
6071 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6072 times from message ids (which are base 36 rather than the normal 62).
6074 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6075 were being counted as actual protocol errors, and logged if the log
6076 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6077 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6078 and DATA following a set of rejected RCPTs do not count as protocol errors.
6079 In other words, Exim assumes they were pipelined, though this may not
6080 actually be the case. Of course, in all cases the client gets an
6081 appropriate error code.
6083 15. If a lookup fails in an ACL condition, a message about the failure may
6084 be available; it is used if testing the ACL cannot continue, because most
6085 such messages specify what the cause of the deferral is. However, some
6086 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6087 that caused an old message to be retained and used if a later statement
6088 caused a defer, replacing the real cause of the deferral.
6090 16. If an IP address had so many PTR records that the DNS lookup buffer
6091 was not large enough to hold them, Exim could crash while trying to process
6092 the truncated data. It now detects and logs this case.
6094 17. Further to 4.21/58, another change has been made: if (and only if) the
6095 first line of a message (the first header line) ends with CRLF, a bare LF
6096 in a subsequent header line has a space inserted after it, so as not to
6097 terminate the header.
6099 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6100 unnecessarily, used atoi() instead of strtol(), and didn't check the
6101 termination when getting file sizes from file names by regex.
6103 19. Completely re-implemented the support for maildirsize files, in the light
6104 of a number of problems with the previous contributed implementation
6105 (4.30/29). In particular:
6107 . If the quota is zero, the maildirsize file is maintained, but no quota is
6110 . If the maildir directory does not exist, it is created before any attempt
6111 to write a maildirsize file.
6113 . The quota value in the file is just a cache; if the quota is changed in
6114 the transport, the new value overrides.
6116 . A regular expression is available for excluding directories from the
6119 20. The autoreply transport checks the characters in options that define the
6120 message's headers; it allows continued headers, but it was checking with
6121 isspace() after an embedded newline instead of explicitly looking for a
6124 21. If all the "regular" hosts to which an address was routed had passed their
6125 expiry times, and had not reached their retry times, the address was
6126 bounced, even if fallback hosts were defined. Now Exim should go on to try
6129 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6130 equivalent code in the SMTP transport. Some hosts send humungous responses
6131 to HELO/EHLO, more than 1024 it seems.
6133 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6134 gives compiler warnings in some environments. I've now done it "properly",
6137 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6138 (because of problems with the built-in one) was declared to return uschar *
6139 instead of char *, causing compiler failure.
6141 25. Fixed a file descriptor leak when processing alias/forward files.
6143 26. Fixed a minor format string issue in dbfn.c.
6145 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6147 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6148 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6151 29. When a sender address is verified, it is cached, to save repeating the test
6152 when there is more than one recipient in a message. However, when the
6153 verification involves a callout, it is possible for different callout
6154 options to be set for different recipients. It is too complicated to keep
6155 track of this in the cache, so now Exim always runs a verification when a
6156 callout is required, relying on the callout cache for the optimization.
6157 The overhead is duplication of the address routing, but this should not be
6160 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6161 to be verified with callout=postmaster, and the main callout worked but the
6162 postmaster check failed, the verification correctly failed. However, if a
6163 subsequent RCPT command asked for sender verification *without* the
6164 postmaster check, incorrect caching caused this verification also to fail,
6167 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6168 it was not caching the DNS options (qualify_single, search_parents) that
6169 were used when the lookup failed. A subsequent lookup with different
6170 options therefore always gave the same answer, though there were cases
6171 where it should not have. (Example: a "domains = !$mx_any" option on a
6172 dnslookup router: the "domains" option is always processed without any
6173 widening, but the router might have qualify_single set.) Now Exim uses the
6174 cached value only when the same options are set.
6176 32. Added John Jetmore's "exipick" utility to the distribution.
6178 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6179 than a timeout (e.g. a certificate is required, and is not provided), an
6180 Exim server now closes the connection immediately. Previously it waited for
6181 the client to close - but if the client is SSL, it seems that they each
6182 wait for each other, leading to a delay before one of them times out.
6184 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6185 maintained 0.8.x compatibility because I don't think many are using it, and
6186 it is clearly obsolete.
6188 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6191 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6192 client certificate was expired. A simple patch fixes this, though I don't
6193 understand the full logic of why the verify callback is called multiple
6196 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6197 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6198 which causes problems with some clients (such as the Certicom SSL Plus
6199 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6200 disables the coutermeasure allowing Eudora to connect."
6202 38. Exim was not checking that a write() to a log file succeeded. This could
6203 lead to Bad Things if a log got too big, in particular if it hit a file
6204 size limit. Exim now panics and dies if it cannot write to a log file, just
6205 as it does if it cannot open a log file.
6207 39. Modified OS/Makefile-Linux so that it now contains
6209 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6211 The two -D definitions ensure that Exim is compiled with large file
6212 support, which makes it possible to handle log files that are bigger than
6215 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6216 instance) a domain was checked against a named list that involved a lookup,
6217 causing $domain_data to be set, then another domain was checked against the
6218 same list, then the first domain was re-checked, the value of $domain_data
6219 after the final check could be wrong. In particular, if the second check
6220 failed, it could be set empty. This bug probably also applied to
6223 41. The strip_trailing_dot option was not being applied to the address given
6224 with the -f command-line option.
6226 42. The code for reading a message's header from the spool was incrementing
6227 $received_count, but never initializing it. This meant that the value was
6228 incorrect (doubled) while delivering a message in the same process in which
6229 it was received. In the most common configuration of Exim, this never
6230 happens - a fresh exec is done - but it can happen when
6231 deliver_drop_privilege is set.
6233 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6234 it now includes up to 150 characters of the unexpected data in the log
6237 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6238 and building data strings. The size of both of these buffers was 10 000
6239 bytes - far larger than anybody would *ever* want, thought I. Needless to
6240 say, somebody hit the limit. I have increased the maximum line length to
6241 20 000 and the maximum data length of concatenated lines to 100 000. I have
6242 also fixed two bugs, because there was no checking on these buffers. Tsk,
6243 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6244 buffer is too small.
6246 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6247 lsearch lookups. Now it does.
6249 46. When parsing a route_list item in a manualroute router, a fixed-length
6250 buffer was used for the list of hosts. I made this 1024 bytes long,
6251 thinking that nobody would ever have a list of hosts that long. Wrong.
6252 Somebody had a whole pile of complicated expansion conditions, and the
6253 string was silently truncated, leading to an expansion error. It turns out
6254 that it is easier to change to an unlimited length (owing to other changes
6255 that have happened since this code was originally written) than to build
6256 structure for giving a limitation error. The length of the item that
6257 expands into the list of hosts is now unlimited.
6259 47. The lsearch lookup could not handle data where the length of text line was
6260 more than 4095 characters. Such lines were truncated, leading to shortened
6261 data being returned. It should now handle lines of any length.
6263 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6264 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6267 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6268 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6269 no longer happens in comment lines. A list of replacements is now placed
6270 at the head of all of the source files, except those whose only change is
6271 to replace PERL_COMMAND in the very first #! line.
6273 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6274 messages on the queue, with a bottom-up merge sort, using code contributed
6275 by Michael Haardt. This should make operations like -bp somewhat faster on
6276 large queues. It won't affect queue runners, except when queue_run_in_order
6279 51. Installed eximstats 1.31 in the distribution.
6281 52. Added support for SRV lookups to the dnslookup router.
6283 53. If an ACL referred to $message_body or $message_body_end, the value was not
6284 reset for any messages that followed in the same SMTP session.
6286 54. The store-handling optimization for building very long strings was not
6287 differentiating between the different store pools. I don't think this
6288 actually made any difference in practice, but I've tidied it.
6290 55. While running the routers to verify a sender address, $sender_address
6291 was still set to the sender address. This is wrong, because when routing to
6292 send a bounce to the sender, it would be empty. Therefore, I have changed
6293 it so that, while verifying a sender address, $sender_address is set to <>.
6294 (There is no change to what happens when verifying a recipient address.)
6296 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6297 target A or AAAA records (if not already returned) without resetting the
6298 qualify_single or search_parents options of the DNS resolver. These are
6299 inappropriate in this case because the targets of MX and SRV records must
6300 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6301 target that, when qualified, matched something in the local domain. These
6302 two options are now turned off when doing these lookups.
6304 57. It seems that at least some releases of Reiserfs (which does not have the
6305 concept of a fixed number of inodes) returns zero and not -1 for the
6306 number of available inodes. This interacted badly with check_spool_inodes,
6307 which assumed that -1 was the "no such thing" setting. What I have done is
6308 to check that the total number of inodes is greater than zero before doing
6309 the test of how many are available.
6311 58. When a "warn" ACL statement has a log_message modifier, the message is
6312 remembered, and not repeated. This is to avoid a lot of repetition when a
6313 message has many recipients that cause the same warning to be written.
6314 However, Exim was preserving the list of already written lines for an
6315 entire SMTP session, which doesn't seem right. The memory is now reset if a
6316 new message is started.
6318 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6319 debugging output unless log_rewrite was also set.
6321 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6322 of (int)(handle) when we know that handle contains (void *)(-1).
6324 61. The Exim daemon panic-logs an error return when it closes the incoming
6325 connection. However "connection reset by peer" seems to be common, and
6326 isn't really an error worthy of noting specially, so that particular error
6329 62. When Exim is trying to find all the local interfaces, it used to panic and
6330 die if the ioctl to get the interface flags failed. However, it seems that
6331 on at least one OS (Solaris 9) it is possible to have an interface that is
6332 included in the list of interfaces, but for which you get a failure error
6333 for this call. This happens when the interface is not "plumbed" into a
6334 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6335 failure of the "get flags" call assumes that the interface is down.
6337 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6338 makes life easier for people who are doing arithmetic on fields extracted
6339 from dates, where you often get leading zeros that should not be
6340 interpreted as octal.
6342 64. Added qualify_domain to the redirect router, to override the global
6345 65. If a pathologically long header line contained very many addresses (the
6346 report of this problem mentioned 10 000) and each of them was rewritten,
6347 Exim could use up a very large amount of memory. (It kept on making new
6348 copies of the header line as it rewrote, and never released the old ones.)
6349 At the expense of a bit more processing, the header rewriting function has
6350 been changed so that it no longer eats memory in this way.
6352 66. The generation of the Received: header has been moved from the time that a
6353 message starts to be received, to the time that it finishes. The timestamp
6354 in the Received: header should now be very close to that of the <= log
6355 line. There are two side-effects of this change:
6357 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6358 logged header lines no longer include the local Received: line, because
6359 it has not yet been created. The same applies to a copy of the message
6360 that is returned to a non-SMTP sender when a message is rejected.
6362 (b) When a filter file is tested using -bf, no additional Received: header
6363 is added to the test message. After some thought, I decided that this
6366 This change does not affect the value of $received_for. It is still set
6367 after address rewriting, but before local_scan() is called.
6369 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6371 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6372 gave an unhelpful panic error message, and a defer error. I have managed to
6373 change this behaviour so that it now rejects any supplied certificate,
6374 which seems right, as the list of acceptable certificates is empty.
6376 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6377 gave an unhelpful defer error. I have not managed to make this reject any
6378 supplied certificates, but the error message it gives is "no certificate
6379 supplied", which is not helpful.
6381 70. exigrep's output now also includes lines that are not associated with any
6382 message, but which match the given pattern. Implemented by a patch from
6383 Martin Sluka, which also tidied up the Perl a bit.
6385 71. Recipient callout verification, like sender verification, was using <> in
6386 the MAIL FROM command. This isn't really the right thing, since the actual
6387 sender may affect whether the remote host accepts the recipient or not. I
6388 have changed it to use the actual sender in the callout; this means that
6389 the cache record is now keyed on a recipient/sender pair, not just the
6390 recipient address. There doesn't seem to be a real danger of callout loops,
6391 since a callout by the remote host to check the sender would use <>.
6392 [SEE ABOVE: changed after hitting problems.]
6394 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6395 temporary errors. However, in the case of such a code being given after
6396 the end of a data transmission (i.e. after ".") Exim was failing to write
6397 a retry record for the message. (Yes, there was some broken host that was
6398 actually sending 8xx at this point.)
6400 73. An unknown lookup type in a host list could cause Exim to panic-die when
6401 the list was checked. (An example that provoked this was putting <; in the
6402 middle of a list instead of at the start.) If this happened during a DATA
6403 ACL check, a -D file could be left lying around. This kind of configuration
6404 error no longer causes Exim to die; instead it causes a defer error. The
6405 incident is still logged to the main and panic logs.
6407 74. Buglet left over from Exim 3 conversion. The message "too many messages
6408 in one connection" was written to the rejectlog but not the mainlog, except
6409 when address rewriting (yes!) was being logged.
6411 75. Added write_rejectlog option.
6413 76. When a system filter was run not as root (that is, when system_filter_user
6414 was set), the values of the $n variables were not being returned to the
6415 main process; thus, they were not subsequently available in the $sn
6418 77. Added +return_path_on_delivery log selector.
6420 78. A connection timeout was being treated differently from recipients deferred
6421 when testing hosts_max_try with a message that was older than the host's
6422 retry timeout. (The host should not be counted, thus allowing all hosts to
6423 be tried at least once before bouncing.) This may have been the cause of an
6424 occasionally reported bug whereby a message would remain on the queue
6425 longer than the retry timeout, but would be bounced if a delivery was
6426 forced. I say "may" because I never totally pinned down the problem;
6427 setting up timeout/retry tests is difficult. See also the next item.
6429 79. The ultimate address timeout was not being applied to errors that involved
6430 a combination of host plus message (for example, a timeout on a MAIL
6431 command). When an address resolved to a number of possible hosts, and they
6432 were not all tried for each delivery (e.g. because of hosts_max_try), a
6433 message could remain on the queue longer than the retry timeout.
6435 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6438 81. Fixed an obscure SMTP outgoing bug which required at least the following
6439 conditions: (a) there was another message waiting for the same server;
6440 (b) the server returned 5xx to all RCPT commands in the first message so
6441 that the message was not completed; (c) the server dropped the connection
6442 or gave a negative response to the RSET that Exim sends to abort the
6443 transaction. The observed case was a dropped connection after DATA that had
6444 been sent in pipelining mode. That is, the server had advertised PIPELINING
6445 but was not implementing it correctly. The effect of the bug was incorrect
6446 behaviour, such as trying another host, and this could lead to a crash.
6452 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6453 and daemon.c were passed as pointers to ints; they should have been
6454 pointers to socklen_t variables (which are typically unsigned ints).
6456 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6459 3. Fixed a really odd bug that affected only the testing scheme; patching a
6460 certain fixed string in the binary changed the value of another string that
6461 happened to be identical to the end of the original first string.
6463 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6464 name", it returns that address as the IP address. On some operating
6465 systems (e.g. Solaris), it also passes back the IP address string as the
6466 "host name". However, on others (e.g. Linux), it passes back an empty
6467 string. Exim wasn't checking for this, and was changing the host name to an
6468 empty string, assuming it had been canonicalized.
6470 5. Although rare, it is permitted to have more than one PTR record for a given
6471 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6472 all the names associated with an address, because they do in Solaris.
6473 However, it seems that they do not in Linux for data that comes from the
6474 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6475 I found this out when I moved to a new Linux workstation and tried to run
6476 the Exim test suite.
6478 To get round this problem I have changed the code so that it now does its
6479 own call to the DNS to look up PTR records when searching for a host name.
6480 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6481 addresses that are only in /etc/hosts are still found.
6483 This behaviour is, however, controlled by an option called host_lookup_
6484 order, which defaults to "bydns:byaddr". If people want to use the other
6485 order, or indeed, just use one or the other means of lookup, they can
6486 specify it in this variable.
6488 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6489 some operating systems, this comes back from gethostbyaddr() as an empty
6490 string, and this is what Exim used to test for. However, it seems that in
6491 other systems, "." is yielded. Exim now tests for this case too.
6493 7. The values of check_spool_space and check_log_space are now held internally
6494 as a number of kilobytes instead of an absolute number of bytes. If a
6495 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6496 kilobyte. This means that much larger values can be stored.
6498 8. Exim monitor: an attempt to get the action menu when not actually pointing
6499 at a message produces an empty menu entitled "No message selected". This
6500 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6501 no entries in it ("Shell widget menu has zero width and/or height"). So I
6502 have added a single, blank menu entry in this case.
6504 9. Added ${quote_local_part.
6506 10. MIME decoding is now applied to the contents of Subject: header lines when
6509 11. Now that a reference to $sender_host_address automatically causes a reverse
6510 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6511 host lookup before query-style lookups in lists that might use this
6512 variable. This has therefore been abolished, and the "net-" prefix is no
6513 longer necessary for query-style lookups.
6515 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6516 have been a typo for LFLAGS, so it has been changed.
6518 13. The install script calls Exim with "-C /dev/null" in order to find the
6519 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6520 to be output. However, since Exim outputs its version number before the
6521 error, it didn't break the script. It just looked ugly. I fixed this by
6522 always allowing "-C /dev/null" if the caller is root.
6524 14. Ignore overlarge ACL variable number when reading spool file - insurance
6525 against a later release with more variables having written the file.
6527 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6528 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6529 this, as well as the form without the "IPv6" on the front.
6531 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6532 OS/Makefile-Darwin file.
6534 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6535 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6537 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6538 variable. However, some OS use ints for inodes. I've added cast to long int
6539 to get rid of the compiler warning.
6541 19. I had forgotten to lock out "/../" in configuration file names when
6542 ALT_CONFIG_PREFIX was set.
6544 20. Routers used for verification do not need to specify transports. However,
6545 if such a router generated a host list, and callout was configured, Exim
6546 crashed, because it could not find a port number from the (non-existent)
6547 transport. It now assumes port 25 in this circumstance.
6549 21. Added the -t option to exigrep.
6551 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6552 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6553 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6554 them is set, LOOKUP_LSEARCH is forced.
6556 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6558 24. Added sender and host information to the "rejected by local_scan()" log
6559 line; previously there was no indication of these.
6561 25. Added .include_if_exists.
6563 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6564 Linux. It turns out that not all file systems support this. Apparently some
6565 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6566 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6567 not supported on the file descriptor, is now ignored when Exim is trying to
6568 sync a directory. This applies only to Linux.
6570 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6572 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6573 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6576 29. Added support for maildirsize files from supplied patch (modified a bit).
6578 30. The use of :fail: followed by an empty string could lead Exim to respond to
6579 sender verification failures with (e.g.):
6581 550 Verification failed for <xxx>
6582 550 Sender verify failed
6584 where the first response line was missing the '-' that indicates it is not
6585 the final line of the response.
6587 31. The loop for finding the name of the user that called Exim had a hardwired
6588 limit of 10; it now uses the value of finduser_retries, which is used for
6589 all other user lookups.
6591 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6594 33. Exim was neglecting to zero errno before one call of strtol() when
6595 expanding a string and expecting an integer value. On some systems this
6596 resulted in spurious "integer overflow" errors. Also, it was casting the
6597 result into an int without checking.
6599 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6600 did not work. The code looks as if it has *never* worked, though it appears
6601 to have been documented since at least release 1.62. I have made it work.
6603 35. The "timeout_DNS" error in retry rules, also documented since at least
6604 1.62, also never worked. As it isn't clear exactly what this means, and
6605 clearly it isn't a major issue, I have abolished the feature by treating it
6606 as "timeout", and writing a warning to the main and panic logs.
6608 36. The display of retry rules for -brt wasn't always showing the error code
6611 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6612 timeout_connect_A, timeout_connect_MX.
6614 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6615 to the empty sender.
6617 39. The daemon was not analysing the content of -oX till after it had closed
6618 stderr and disconnected from the controlling terminal. This meant that any
6619 syntax errors were only noted on the panic log, and the return code from
6620 the command was 0. By re-arranging the code a little, I've made the
6621 decoding happen first, so such errors now appear on stderr, and the return
6622 code is 1. However, the actual setting up of the sockets still happens in
6623 the disconnected process, so errors there are still only recorded on the
6626 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6627 connections (as happens on some IP stacks) was logged at start up time as
6628 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6629 it from "IPv6 and IPv4", which means that two separate sockets are being
6632 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6633 says whether AF_INET or AF_INET6 was passed as an argument.
6635 42. Exiwhat output was messed up when time zones were included in log
6638 43. Exiwhat now gives more information about the daemon's listening ports,
6639 and whether -tls-on-connect was used.
6641 44. The "port" option of the smtp transport is now expanded.
6643 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6644 silently ignored. Now an error message is written to the main and panic
6647 46. There's a new ACL modifier called "logwrite" which writes to a log file
6648 as soon as it is encountered.
6650 47. Added $local_user_uid and $local_user_gid at routing time.
6652 48. Exim crashed when trying to verify a sender address that was being
6655 49. Exim was recognizing only a space character after ".include". It now also
6656 recognizes a tab character.
6658 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6659 extracting the relevant information from the specification. The man page no
6660 longer contains scrambled data for the -d option, and I've added a section
6661 at the front about calling Exim under different names.
6663 51. Added "extra_headers" argument to the "mail" command in filter files.
6665 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6668 53. Installed eximstats 1.29.
6670 54. Added transport_filter_timeout as a generic transport option.
6672 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6673 Cc: header lines. This was required by RFC 822, but it not required by RFC
6676 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6677 incoming messages that did not have them. Now it does so only if the
6678 message originates locally, that is, if there is no associated remote host
6679 address. When Resent- header lines are present, this applies to the Resent-
6680 lines rather than the non-Resent- lines.
6682 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6683 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6685 58. Messages for configuration errors now include the name of the main
6686 configuration file - useful now that there may be more than one file in a
6687 list (.included file names were always shown).
6689 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6690 for those rare installations that do not start the daemon as root or run it
6691 setuid root. I've cut out the call to initgroups() if the daemon is not
6694 60. The Exim user and group can now be bound into the binary as text strings
6695 that are looked up at the start of Exim's processing.
6697 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6699 62. Added $mailstore_basename variable.
6701 63. Installed patch to sieve.c from Michael Haardt.
6703 64. When Exim failed to open the panic log after failing to open the main log,
6704 the original message it was trying to log was written to stderr and debug
6705 output, but if they were not available (the usual case in production), it
6706 was lost. Now it is written to syslog before the two lines that record the
6707 failures to open the logs.
6709 65. Users' Exim filters run in subprocesses under the user's uid. It is
6710 possible for a "deliver" command or an alias in a "personal" command to
6711 provoke an address rewrite. If logging of address rewriting is configured,
6712 this fails because the process is not running as root or exim. There may be
6713 a better way of dealing with this, but for the moment (because 4.30 needs
6714 to be released), I have disabled address rewrite logging when running a
6715 filter in a non-root, non-exim process.
6721 1. The buildconfig auxiliary program wasn't quoting the value set for
6722 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6723 not defined. This bug was masked in 4.22 by the effect that was fixed in
6726 2. Some messages that were rejected after a message id was allocated were
6727 shown as "incomplete" by exigrep. It no longer does this for messages that
6728 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6730 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6731 have allow_domain_literals set, the ID did not get logged in the <= line.
6732 Domain literals are now always recognized in Message-ID: header lines.
6734 4. The first argument for a ${extract expansion item is the key name or field
6735 number. Leading and trailing spaces in this item were not being ignored,
6736 causing some misleading effects.
6738 5. When deliver_drop_privilege was set, single queue runner processes started
6739 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6740 same command in the process it spins off) were not dropping privilege.
6742 6. When the daemon running as "exim" started a queue runner, it always
6743 re-executed Exim in the spun-off process. This is a waste of effort when
6744 deliver_drop_privilege is set. The new process now just calls the
6745 queue-runner function directly.
6751 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6754 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6755 ignored. Though the use of -oP was forcing the writing of a pid file, it
6756 was always written to the default place.
6758 3. If the message "no IP address found for host xxxx" is generated during
6759 incoming verification, it is now followed by identification of the incoming
6760 connection (so you can more easily find what provoked it).
6762 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6764 5. Added some features to "harden" Exim a bit more against certain attacks:
6766 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6767 be put in Local/Makefile. This is like the never_users runtime option,
6768 but it cannot be overridden. The default setting is "root".
6770 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6771 prefix string with which any file named in a -C command line option
6774 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6775 is retained for -C and -D only if the caller of Exim is root. Without
6776 it, the exim user may also use -C and -D and retain privilege.
6778 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6779 command line option is disabled.
6781 6. Macro names set by the -D option must start with an upper case letter, just
6782 like macro names defined in the configuration file.
6784 7. Added "dereference=" facility to LDAP.
6786 8. Two instances of the typo "uknown" in the source files are fixed.
6788 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6789 the Configure-Makefile script screwed up while processing it.
6791 10. Incorporated PCRE 4.4.
6793 11. The SMTP synchronization check was not operating right at the start of an
6794 SMTP session. For example, it could not catch a HELO sent before the client
6795 waited for the greeting. There is now a check for outstanding input at the
6796 point when the greeting is written. Because of the duplex, asynchronous
6797 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6798 way, but not yet received, when the check is performed.
6800 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6801 on TCP/IP sockets, because this apparently causes some broken clients to
6804 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6805 unchanged) from the Cygwin maintainer.
6807 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6808 was supported instead of testing for mbx. Effectively a typo.
6810 15. The spa authenticator server code was not checking that the input it
6811 received was valid base64.
6813 16. The debug output line for the "set" modifier in ACLs was not showing the
6814 name of the variable that was being set.
6816 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6818 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6819 lookup on its own. Something else was needed to trigger the lookup. For
6820 example, a match in host_lookup or the need for a host name in a host list.
6821 Now, if $sender_host_name is referenced and the host name has not yet been
6822 looked up, a lookup is performed. If the lookup fails, the variable remains
6823 empty, and $host_lookup_failed is set to "1".
6825 19. Added "eqi" as a case-independent comparison operator.
6827 20. The saslauthd authentication condition could segfault if neither service
6828 nor realm was specified.
6830 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6831 error message that was logged was misleading, and incoming SMTP
6832 connections were dropped. The message is now more accurate, and temporary
6833 errors are given to SMTP connections.
6835 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6836 (except RSET) with a 421 error, until QUIT is received. However, it was
6837 failing to send a response to QUIT.
6839 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6840 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6841 ACL is now run after the helo_try_verify_hosts code.
6843 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6844 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6845 case-independent, so other case variants are also recognized). Apparently
6846 some systems use these upper case variants.
6848 25. If more than two messages were waiting for the same host, and a transport
6849 filter was specified for the transport, Exim sent two messages over the
6850 same TCP/IP connection, and then failed with "socket operation on non-
6851 socket" when it tried to send the third.
6853 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6855 27. The extern definition of crypt16() in expand.c was not being excluded when
6856 the OS had its own crypt16() function.
6858 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6859 as a preferred synonym for return_size_limit, both as an option and as an
6862 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6864 30. Changed the default configuration ACL to relax the local part checking rule
6865 for addresses that are not in any local domains. For these addresses,
6866 slashes and pipe symbols are allowed within local parts, but the sequence
6867 /../ is explicitly forbidden.
6869 31. SPA server authentication was not clearing the challenge buffer before
6872 32. log_message in a "warn" ACL statement was writing to the reject log as
6873 well as to the main log, which contradicts the documentation and doesn't
6874 seem right (because no rejection is happening). So I have stopped it.
6876 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6877 However, I am unable to do any testing of this.
6879 34. Fixed an infelicity in the appendfile transport. When checking directories
6880 for a mailbox, to see if any needed to be created, it was accidentally
6881 using path names with one or more superfluous leading slashes; tracing
6882 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6884 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6885 contents are added to the log line that is written for every discarded
6886 recipient. (Previously a log_message setting was ignored.)
6888 36. The ${quote: operator now quotes the string if it is empty.
6890 37. The install script runs exim in order to find its version number. If for
6891 some reason other than non-existence or emptiness, which it checks, it
6892 could not run './exim', it was installing it with an empty version number,
6893 i.e. as "exim-". This error state is now caught, and the installation is
6896 38. An argument was missing from the function that creates an error message
6897 when Exim fails to connect to the socket for saslauthd authentication.
6898 This could cause Exim to crash, or give a corrupted message.
6900 39. Added isip, isip4, and isip6 to ${if conditions.
6902 40. The ACL variables $acl_xx are now saved with the message, and can be
6903 accessed later in routers, transports, and filters.
6905 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6906 strings in the file are not string-expanded.
6908 42. If a MAIL command specified a SIZE value that was too large to fit into an
6909 int variable, the check against message_size_limit failed. Such values are
6910 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6911 day this will have to be increased, but I don't think I want to be around
6912 when emails are that large.
6919 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6920 iconv() is not standard in FreeBSD.
6922 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6923 IPv6 enabled. The observed symptom was a segmentation fault on return from
6924 the function os_common_find_running_interfaces() in src/os.c.
6926 3. In the check_special_case() function in daemon.c I had used "errno" as an
6927 argument name, which causes warnings on some systems. This was basically a
6928 typo, since it was named "eno" in the comments!
6930 4. The code that waits for the clock to tick (at a resolution of some fraction
6931 of a second) so as to ensure message-id uniqueness was always waiting for
6932 at least one whole tick, when it could have waited for less. [This is
6933 almost certainly not relevant at current processor speeds, where it is
6934 unlikely to ever wait at all. But we try to future-proof.]
6936 5. The function that sleeps for a time interval that includes fractions of a
6937 second contained a race. It did not block SIGALRM between setting the
6938 timer, and suspending (a couple of lines later). If the interval was short
6939 and the sigsuspend() was delayed until after it had expired, the suspension
6940 never ended. On busy systems this could lead to processes getting stuck for
6943 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6944 process, before it forks any delivery processes. The open lookup caching
6945 mechanism meant that the open file or database connection was passed into
6946 the delivery process. The problem was that delivery processes always tidy
6947 up cached lookup data. This could cause a problem for the next delivery
6948 process started by the queue runner, because the external queue runner
6949 process does not know about the closure. So the next delivery process
6950 still has data in the lookup cache. In the case of a file lookup, there was
6951 no problem because closing a file descriptor in a subprocess doesn't affect
6952 the parent. However, if the lookup was caching a connection to a database,
6953 the connection was closed, and the second delivery process was likely to
6954 see errors such as "PGSQL: query failed: server closed the connection
6955 unexpectedly". The problem has been fixed by closing all cached lookups
6956 in a queue runner before running a delivery process.
6958 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6959 seem to have the "const" qualifier which it has on other OS. I've
6962 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6963 *of the same type* (client or server) with the same public name that an
6964 error should be diagnosed.
6966 9. When Exim looked up a host name for an IP address, but failed to find the
6967 original IP address when looking up the host name (a safety check), it
6968 output the message "<ip address> does not match any IP for NULL", which was
6969 confusing, to say the least. The bug was that the host name should have
6970 appeared instead of "NULL".
6972 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6973 user that is built into the binary, and the -C or -D options is used, root
6974 privilege is dropped before the configuration file is read. In addition,
6975 logging is switched to stderr instead of the normal log files. If the
6976 configuration then re-defines the Exim user, the unprivileged environment
6977 is probably not what is expected, so Exim logs a panic warning message (but
6980 However, if deliver_drop_privilege is set, the unprivileged state may well
6981 be exactly what is intended, so the warning has been cut out in that case,
6982 and Exim is allowed to try to write to its normal log files.
6988 1. smtp_return_error_details was not giving details for temporary sender
6989 or receiver verification errors.
6991 2. Diagnose a configuration error if two authenticators have the same public
6994 3. Exim used not to create the message log file for a message until the first
6995 delivery attempt. This could be confusing when incoming messages were held
6996 for policy or load reasons. The message log file is now created at the time
6997 the message is received, and an initial "Received" line is written to it.
6999 4. The automatically generated man page for command line options had a minor
7000 bug that caused no ill effects; however, a more serious problem was that
7001 the procedure for building the man page automatically didn't always
7002 operate. Consequently, release 4.20 contains an out-of-date version. This
7003 shouldn't happen again.
7005 5. When building Exim with embedded Perl support, the script that builds the
7006 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7007 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7009 6. The freeze_tell option was not being used for messages that were frozen on
7010 arrival, either by an ACL or by local_scan().
7012 7. Added the smtp_incomplete_transaction log selector.
7014 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7015 was accepting AUTH without a new EHLO.
7017 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7018 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7021 10. Logging of TCP/IP connections (when configured) now happens in the main
7022 daemon process instead of the child process, so that the TCP/IP connection
7023 count is more accurate (but it can never be perfect).
7025 11. The use of "drop" in a nested ACL was not being handled correctly in the
7026 outer ACL. Now, if condition failure induced by the nested "drop" causes
7027 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7028 or "require"), the connection is dropped.
7030 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7031 that yield "discard" can now be used with an "accept" or a "discard" verb,
7032 but an error is generated for any others (because I can't see a useful way
7033 to define what should happen).
7035 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7036 are now processed in the same way as lines in the Exim configuration file.
7037 In particular, continuation lines are supported.
7039 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7041 15. Added -ti meaning -t -i.
7043 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7044 domains, and warn by logging if others are found.
7046 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7047 returned from GIFCONF when Exim is trying to find the list of interfaces on
7048 a host. The code in os.c has been modified to copy each ifreq to an aligned
7049 structure in all cases.
7051 Also, in some cases, the returned ifreq's were being copied to a 'struct
7052 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7053 means the last couple of bytes of an IPv6 address could be chopped if the
7054 ifreq contained only a normal sockaddr (14 bytes storage).
7056 18. Named domain lists were not supported in the hosts_treat_as_local option.
7057 An entry such as +xxxx was not recognized, and was treated as a literal
7060 19. Ensure that header lines added by a DATA ACL are included in the reject log
7061 if the ACL subsequently rejects the message.
7063 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7064 MD5 (which is deprecated).
7066 21. When testing a filter file using -bf, Exim was writing a message when it
7067 took the sender from a "From " line in the message, but it was not doing so
7068 when it took $return_path from a Return-Path: header line. It now does.
7070 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7071 with a valid header line field name (a series of printing characters
7072 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7074 23. Changed "disc" in the source to "disk" to conform to the documentation and
7075 the book and for uniformity.
7077 24. Ignore Sendmail's -Ooption=value command line item.
7079 25. When execve() failed while trying to run a command in a pipe transport,
7080 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7081 could be confused with a return value of 69 from the command itself. This
7082 has been changed to 127, the value the shell returns if it is asked to run
7083 a non-existent command. The wording for the related log line suggests a
7084 non-existent command as the problem.
7086 26. If received_header_text expands to an empty string, do not add a Received:
7087 header line to the message. (Well, it adds a token one on the spool, but
7088 marks it "old" so that it doesn't get used or transmitted.)
7090 27. Installed eximstats 1.28 (addition of -nt option).
7092 28. There was no check for failure on the call to getsockname() in the daemon
7093 code. This can fail if there is a shortage of resources on the system, with
7094 ENOMEM, for example. A temporary error is now given on failure.
7096 29. Contrary to the C standard, it seems that in some environments, the
7097 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7098 program. Exim now does this explicitly; it affects the formatting of
7099 timestamps using strftime().
7101 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7102 complaints. I've now initialized all the variables, to avoid this.
7104 32. Header lines added by a system filter were not being "seen" during
7105 transport-time rewrites.
7107 33. The info_callback() function passed to OpenSSL is set up with type void
7108 (*)(SSL *, int, int), as described somewhere. However, when calling the
7109 function (actually a macro) that sets it up, the type void(*)() is
7110 expected. I've put in a cast to prevent warnings from picky compilers.
7112 34. If a DNS black list lookup found a CNAME record, but there were no A
7113 records associated with the domain it pointed at, Exim crashed.
7115 35. If a DNS black list lookup returned more than one A record, Exim ignored
7116 all but the first. It now scans all returned addresses if a particular IP
7117 value is being sought. In this situation, the contents of the
7118 $dnslist_value variable are a list of all the addresses, separated by a
7121 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7122 to accept a host name and all its aliases if the forward lookup for any of
7123 them yielded the IP address of the incoming connection. Now it accepts only
7124 those names whose forward lookup yields the correct IP address. Any other
7125 names are discarded. This closes a loophole whereby a rogue DNS
7126 administrator could create reverse DNS records to break through a
7127 wildcarded host restriction in an ACL.
7129 37. If a user filter or a system filter that ran in a subprocess used any of
7130 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7131 the wrong values were passed to the pipe command ($thisaddress had the
7132 value of $0, $0 had the value of $1, etc). This bug was introduced by
7133 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7135 38. Improved the line breaking for long SMTP error messages from ACLs.
7136 Previously, if there was no break point between 40 and 75 characters, Exim
7137 left the rest of the message alone. Two changes have been made: (a) I've
7138 reduced the minimum length to 35 characters; (b) if it can't find a break
7139 point between 35 and 75 characters, it looks ahead and uses the first one
7140 that it finds. This may give the occasional overlong line, but at least the
7141 remaining text gets split now.
7143 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7144 file descriptors might be low, and that setting 1000 would always raise it.
7145 It turns out that in some environments, the limit is already over 1000 and
7146 that lowering it causes trouble. So now Exim takes care not to decrease it.
7148 40. When delivering a message, the value of $return_path is set to $sender_
7149 address at the start of routing (routers may change the value). By an
7150 oversight, this default was not being set up when an address was tested by
7151 -bt or -bv, which affected the outcome if any router or filter referred to
7154 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7155 log only when "message" or "log_message" are set. However, if one of the
7156 conditions was an address verification, or a call to a nested ACL, the
7157 messages generated by the underlying test were being passed through. This
7158 no longer happens. The underlying message is available in $acl_verify_
7159 message for both "message" and "log_message" expansions, so it can be
7160 passed through if needed.
7162 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7163 new expansion $bh_ to give the encoded byte string without charset
7164 translation. Translation happens only if iconv() is available; HAVE_ICONV
7165 indicates this at build time. HEADERS_CHARSET gives the charset to
7166 translate to; headers_charset can change it in the configuration, and
7167 "headers charset" can change it in an individual filter file.
7169 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7170 that creates RFC 2047 encoded "words" labels them as that charset instead
7171 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7172 expansion operator; (ii) when Exim creates a From: line for a local
7173 message; (iii) when a header line is rewritten to include a "phrase" part.
7175 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7176 buggy, causing it to skip the first lines of messages whose message ID
7177 ended in 'D'. This would not have bitten before Exim release 4.14, because
7178 message IDs were unlikely to end in 'D' before then. The effect was to have
7179 incorrect size information for certain domains.
7181 45. #include "config.h" was missing at the start of the crypt16.c module. This
7182 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7185 46. If there was a timeout during a "random" callout check, Exim treated it as
7186 a failure of the random address, and carried on sending RSET and the real
7187 address. If the delay was just some slowness somewhere, the response to the
7188 original RCPT would be taken as a response to RSET and so on, causing
7189 mayhem of various kinds.
7191 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7192 when I implemented it. It didn't allow for the fact that some option values
7193 may legitimately be negative (e.g. size_addition), and it didn't even do
7194 the right test for positive values.
7196 48. Domain names in DNS records are case-independent. Exim always looks them up
7197 in lower case. Some resolvers return domain names in exactly the case they
7198 appear in the zone file, that is, they may contain uppercase letters. Not
7199 all resolvers do this - some return always lower case. Exim was treating a
7200 change of case by a resolver as a change of domain, similar to a widening
7201 of a domain abbreviation. This triggered its re-routing code and so it was
7202 trying to route what was effectively the same domain again. This normally
7203 caused routing to fail (because the router wouldn't handle the domain
7204 twice). Now Exim checks for this case specially, and just changes the
7205 casing of the domain that it ultimately uses when it transmits the message
7208 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7211 50. If a filter generated a file delivery with a non-absolute name (possible if
7212 no home directory exists for the router), the forbid_file option was not
7215 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7216 the existing equality matching.
7218 52. Exim was using ints instead of ino_t variables in some places where it was
7219 dealing with inode numbers.
7221 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7222 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7223 called TMPDIR, and if it finds it is different, it changes its value.
7225 54. The smtp_printf() function is now made available to local_scan() so
7226 additional output lines can be written before returning. There is also an
7227 smtp_fflush() function to enable the detection of a dropped connection.
7228 The variables smtp_input and smtp_batched_input are exported to
7231 55. Changed the default runtime configuration: the message "Unknown user"
7232 has been removed from the ACL, and instead placed on the localuser router,
7233 using the cannot_route_message feature. This means that any verification
7234 failures that generate their own messages won't get overridden. Similarly,
7235 the "Unrouteable address" message that was in the ACL for unverifiable
7236 relay addresses has also been removed.
7238 56. Added hosts_avoid_esmtp to the smtp transport.
7240 57. The exicyclog script was not checking for the esoteric option
7241 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7242 will work only if exicyclog is run under the appropriate euid.
7244 58. Following a discussion on the list, the rules by which Exim recognises line
7245 endings on incoming messages have been changed. The -dropcr and drop_cr
7246 options are now no-ops, retained only for backwards compatibility. The
7247 following line terminators are recognized: LF CRLF CR. However, special
7248 processing applies to CR:
7250 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7251 nor a local message in the state where . is a terminator.
7253 (ii) If a bare CR is encountered in a header line, an extra space is added
7254 after the line terminator so as not to end the header. The reasoning
7255 behind this is that bare CRs in header lines are most likely either
7256 to be mistakes, or people trying to play silly games.
7258 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7259 was being incorrectly given as 18 bytes larger than it should have been.
7260 This is a VOB (very old bug).
7262 60. This may never have affected anything current, but just in case it has:
7263 When the local host is found other than at the start of a list of hosts,
7264 the local host, those with the same MX, and any that follow, are discarded.
7265 When the list in question was part of a longer list of hosts, the following
7266 hosts (not currently being processed) were also being discarded. This no
7267 longer happens. I'm not sure if this situation could ever has previously
7270 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7273 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7274 header. This is something that is recommended in a new Internet Draft, and
7275 is something that is documented as being done by Sendmail. There are two
7276 possible values. For messages generated by the autoreply transport, Exim
7279 Auto-Submitted: auto-replied
7281 whereas for all other generated messages (e.g. bounces) it adds
7283 Auto-Submitted: auto-generated
7285 63. The "personal" condition in filters now includes a test for the
7286 Auto-Submitted: header. If it contains the string "auto-" the message it
7287 not considered personal.
7289 64. Added rcpt_include_affixes as a generic transport option.
7291 65. Added queue_only_override (default true).
7293 66. Added the syslog_duplication option.
7295 67. If what should have been the first header line of a message consisted of
7296 a space followed by a colon, Exim was mis-interpreting it as a header line.
7297 It isn't of course - it is syntactically invalid and should therefore be
7298 treated as the start of the message body. The misbehaviour could have
7299 caused a number of strange effects, including loss of data in subsequent
7300 header lines, and spool format errors.
7302 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7303 client host had authenticated. This control can now be exercised by an ACL
7304 for more flexibility.
7306 69. By default, callouts do not happen when testing with -bh. There is now a
7307 variant, -bhc, which does actually run the callout code, including
7308 consulting and updating the callout cache.
7310 70. Added support for saslauthd authentication, courtesy of Alexander
7313 71. If statvfs() failed on the spool or log directories while checking their
7314 size for availability, Exim confusingly gave the error "space shortage".
7315 Furthermore, in debugging mode it crashed with a floating point exception.
7316 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7317 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7318 serious problem, Exim now writes to the main and panic logs when this
7319 happens, with details of the failure. It then refuses to accept the
7320 incoming message, giving the message "spool directory problem" or "log
7321 directory problem" with a 421 code for SMTP messages.
7323 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7324 0, 1, and 2 exist, because some OS complain for execs without them (see
7325 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7326 descriptors. However, the code omitted to check that the open succeeded,
7327 causing mysterious errors if for some reason the permissions on /dev/null
7328 got screwed. Now Exim writes a message to the main and panic logs, and
7329 bombs out if it can't open /dev/null.
7331 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7332 interact so that it is all more flexible. It is supposed to remain
7333 backwards compatible. Also added extra_local_interfaces.
7335 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7336 to bomb out with an assertion failure - to the client this appears as a
7337 connection drop. This problem occurs in the part of the code that was taken
7338 from the Samba project. Fortunately, the assertion is in a very simple
7339 function, so I have fixed this by reproducing the function inline in the
7340 one place where it is called, and arranging for authentication to fail
7341 instead of killing the process with assert().
7343 75. The SPA client code was not working when the server requested OEM rather
7344 than Unicode encoding.
7346 76. Added code to make require_files with a specific uid setting more usable in
7347 the case where statting the file as root fails - usually a non-root-mounted
7348 NFS file system. When this happens and the failure is EACCES, Exim now
7349 forks a subprocess and does the per-uid checking as the relevant uid.
7351 77. Added process_log_path.
7353 78. If log_file_path was not explicitly set, a setting of check_log_space or
7354 check_log_inodes was ignored.
7356 79. If a space check for the spool or log partitions fails, the incident is now
7357 logged. Of course, in the latter case the data may get lost...
7359 80. Added the %p formatting code to string_format() so that it can be used to
7360 print addresses in debug_print(). Adjusted all the address printing in the
7361 debugging in store.c to use %p rather than %d.
7363 81. There was a concern that a line of code in smtp_in.c could overflow a
7364 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7365 initially expressed, the concern was not well-founded, because trailing
7366 spaces are removed early. However, if the trailing spaces were followed by
7367 a NULL, they did not get removed, so the overflow was possible. Two fixes
7370 (a) I re-wrote the offending code in a cleaner fashion.
7371 (b) If an incoming SMTP command contains a NULL character, it is rejected
7374 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7375 runs initgroups(), so that if the Exim user is in any additional groups,
7376 they will be used during message reception.
7382 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob