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 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually removed using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
119 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
120 a "Gnu special" function, asprintf() in the DB utility binary builds; I
121 hope that is portable enough.
123 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
124 requiring a known-CA anchor certificate; make it now rely entirely on the
125 TLSA as an anchor. Checking the name on the leaf cert against the name
126 on the A-record for the host is still done for TA (but not for EE mode).
128 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
129 included in delivery lines for non-proxied connections, when compiled with
130 SUPPORT_SOCKS and running with proxy logging enabled.
132 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
133 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
134 move the existing event to fire before the normal logging of message
135 failure so that custom logging is bracketed by normal logging.
141 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
142 When redis_servers is set to a list of > 1 element, and the Redis servers
143 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
144 case of MOVED into a DEFER case instead, thus moving the query onto the
145 next server in the list. For a cluster of N elements, all N servers must
146 be defined in redis_servers.
148 GF/02 Catch and remove uninitialized value warning in exiqsumm
149 Check for existence of @ARGV before looking at $ARGV[0]
151 JH/01 Replace the store_release() internal interface with store_newblock(),
152 which internalises the check required to safely use the old one, plus
153 the allocate and data copy operations duplicated in both (!) of the
154 extant use locations.
156 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
157 modifier. This matches the restriction on the commandline.
159 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
160 Previously only the last row was returned.
162 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
163 we assumed that tags in the header were well-formed, and parsed the
164 element content after inspecting only the first char of the tag.
165 Assumptions at that stage could crash the receive process on malformed
168 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
169 While running the DKIM ACL we operate on the Permanent memory pool so that
170 variables created with "set" persist to the DATA ACL. Also (at any time)
171 DNS lookups that fail create cache records using the Permanent pool. But
172 expansions release any allocations made on the current pool - so a dnsdb
173 lookup expansion done in the DKIM ACL releases the memory used for the
174 DNS negative-cache, and bad things result. Solution is to switch to the
175 Main pool for expansions.
176 While we're in that code, add checks on the DNS cache during store_reset,
177 active in the testsuite.
178 Problem spotted, and debugging aided, by Wolfgang Breyha.
180 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
181 When none of the hosts presented to a transport match an already-open
182 connection, close it and proceed with the list. Previously we would
183 queue the message. Spotted by Lena with Yahoo, probably involving
186 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
187 Previously a spurious "250 OK id=" response was appended to the proper
190 JH/08 The "support for" informational output now, which built with Content
191 Scanning support, has a line for the malware scanner interfaces compiled
192 in. Interface can be individually included or not at build time.
194 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
195 by the template makefile "src/EDITME". The "STREAM" support for an older
196 ClamAV interface method is removed.
198 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
199 rows affected is given instead).
201 JH/11 The runtime Berkeley DB library version is now additionally output by
202 "exim -d -bV". Previously only the compile-time version was shown.
204 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
205 SMTP connection. Previously, when one had more receipients than the
206 first, an abortive onward connection was made. Move to full support for
207 multiple onward connections in sequence, handling cutthrough connection
208 for all multi-message initiating connections.
210 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
211 routers. Previously, a multi-recipient message would fail to match the
212 onward-connection opened for the first recipient, and cause its closure.
214 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
215 a timeout on read on a GnuTLS initiating connection, resulting in the
216 initiating connection being dropped. This mattered most when the callout
217 was marked defer_ok. Fix to keep the two timeout-detection methods
220 JH/15 Relax results from ACL control request to enable cutthrough, in
221 unsupported situations, from error to silently (except under debug)
222 ignoring. This covers use with PRDR, frozen messages, queue-only and
225 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
227 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
228 metadata, resulting in a crash in free().
230 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
231 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
232 Broken also in d185889f4, with init system revamp.
234 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
235 Previously we abruptly closed the connection after reading a malware-
236 found indication; now we go on to read the "scan ok" response line,
239 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
240 ACL. Previously, a crash would result.
242 JH/19 Speed up macro lookups during configuration file read, by skipping non-
243 macro text after a replacement (previously it was only once per line) and
244 by skipping builtin macros when searching for an uppercase lead character.
246 JH/20 DANE support moved from Experimental to mainline. The Makefile control
247 for the build is renamed.
249 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
250 was allocated for every new TLS startup, meaning one per message. Fix
251 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
253 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
254 reported the original. Fix to report (as far as possible) the ACL
255 result replacing the original.
257 JH/23 Fix memory leak during multi-message connections using STARTTLS under
258 OpenSSL. Certificate information is loaded for every new TLS startup,
259 and the resources needed to be freed.
261 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
263 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
266 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
267 DATA response info to the (existing) per-recipient response info for
268 the "C=" log element. It can have useful tracking info from the
269 destination system. Patch from Simon Arlott.
271 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
272 length value. Previously this would segfault.
274 HS/02 Support Avast multiline protoocol, this allows passing flags to
275 newer versions of the scanner.
277 JH/28 Ensure that variables possibly set during message acceptance are marked
278 dead before release of memory in the daemon loop. This stops complaints
279 about them when the debug_store option is enabled. Discovered specifically
280 for sender_rate_period, but applies to a whole set of variables.
281 Do the same for the queue-runner and queue-list loops, for variables set
282 from spool message files. Do the same for the SMTP per-message loop, for
283 certain variables indirectly set in ACL operations.
285 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
286 as a multi-recipient message from a mailinglist manager). The coding had
287 an arbitrary cutoff number of characters while checking for more input;
288 enforced by writing a NUL into the buffer. This corrupted long / fast
289 input. The problem was exposed more widely when more pipelineing of SMTP
290 responses was introduced, and one Exim system was feeding another.
291 The symptom is log complaints of SMTP syntax error (NUL chars) on the
292 receiving system, and refused recipients seen by the sending system
293 (propating to people being dropped from mailing lists).
294 Discovered and pinpointed by David Carter.
296 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
297 replaced by the ${authresults } expansion.
299 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
301 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
302 allows proper process termination in container environments.
304 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
305 Previously the "final dot" had a newline after it; ensure it is CR,LF.
307 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
308 and "err_perm", deprecated since 4.83 when the RFC-defined words
309 "temperror" and "permerror" were introduced.
311 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
312 transport-filters or DKIM-signing. The restriction was lost in the
313 consolidation of verify-callout and delivery SMTP handling.
314 Extend the restriction to also cover ARC-signing.
316 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
317 in defer=pass mode supply a 450 to the initiator. Previously the message
320 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
321 tls_require_ciphers is used as before.
323 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
324 "pass_unscanned". Only tmpfails from the scanner are written to
325 the paniclog, as they may require admin intervention (permission
326 denied, license issues). Other scanner errors (like decompression
327 bombs) do not cause a paniclog entry.
329 JH/36 Fix reinitialisation of DKIM logging variable between messages.
330 Previously it was possible to log spurious information in receive log
333 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
334 triggered odd behaviour from Outlook Express clients.
336 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
339 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
340 since the IETF WG has not yet settled on that versus the original
341 "bare" representation.
343 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
344 Previously the millisecond value corrupted the output.
345 Fix also for syslog_pid=no and log_selector +pid, for which the pid
346 corrupted the output.
352 JH/01 Rework error string handling in TLS interface so that the caller in
353 more cases is responsible for logging. This permits library-sourced
354 string to be attached to addresses during delivery, and collapses
355 pairs of long lines into single ones.
357 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
358 during configuration. Wildcards are allowed and expanded.
360 JH/02 Rework error string handling in DKIM to pass more info back to callers.
361 This permits better logging.
363 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
364 do not close it down and have the child transport start it up again on
365 the passed-on TCP connection. Instead, proxy the child (and any
366 subsequent ones) for TLS via a unix-domain socket channel. Logging is
367 affected: the continued delivery log lines do not have any DNSSEC, TLS
368 Certificate or OCSP information. TLS cipher information is still logged.
370 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
371 identical IP addresses on different listening ports. Will also affect
374 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
375 add noisy ifdef guards to special-case this sillyness.
376 Patch from Bernd Kuhls.
378 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
379 than 255 are no longer allowed.
381 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
382 Disable the session-cache too, which might reduce our load. Since we
383 currrectly use a new context for every connection, both as server and
384 client, there is no benefit for these.
385 GnuTLS appears to not support tickets server-side by default (we don't
386 call gnutls_session_ticket_enable_server()) but client side is enabled
387 by default on recent versions (3.1.3 +) unless the PFS priority string
390 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
391 <https://reproducible-builds.org/specs/source-date-epoch/>.
393 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
394 the check for any unsuccessful recipients did not notice the limit, and
395 erroneously found still-pending ones.
397 JH/08 Pipeline CHUNKING command and data together, on kernels that support
398 MSG_MORE. Only in-clear (not on TLS connections).
400 JH/09 Avoid using a temporary file during transport using dkim. Unless a
401 transport-filter is involved we can buffer the headers in memory for
402 creating the signature, and read the spool data file once for the
403 signature and again for transmission.
405 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
406 4.77 as the kernel support then wasn't solid, having issues in 64bit
407 mode. Now, it's been long enough. Add support for FreeBSD also.
409 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
410 case where the routing stage had gathered several addresses to send to
411 a host before calling the transport for the first, we previously failed
412 to close down TLS in the old transport process before passing the TCP
413 connection to the new process. The new one sent a STARTTLS command
414 which naturally failed, giving a failed delivery and bloating the retry
415 database. Investigation and fix prototype from Wolfgang Breyha.
417 JH/12 Fix check on SMTP command input synchronisation. Previously there were
418 false-negatives in the check that the sender had not preempted a response
419 or prompt from Exim (running as a server), due to that code's lack of
420 awareness of the SMTP input buffering.
422 PP/04 Add commandline_checks_require_admin option.
423 Exim drops privileges sanely, various checks such as -be aren't a
424 security problem, as long as you trust local users with access to their
425 own account. When invoked by services which pass untrusted data to
426 Exim, this might be an issue. Set this option in main configuration
427 AND make fixes to the calling application, such as using `--` to stop
430 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
431 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
432 responses to those, into a single TLS record each way (this usually means
433 a single packet). As a side issue, smtp_enforce_sync now works on TLS
436 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
437 affects you only if you're dancing at the edge of the param size limits.
438 If you are, and this message makes sense to you, then: raise the
439 configured limit or use OpenSSL 1.1. Nothing we can do for older
442 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
443 cmdline element to get the documented default one. Previously it was
446 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
447 are unsuable for non-message-specific reasons - eg. connection timeout,
448 banner-time rejection.
450 JH/16 Fix logging of delivery remote port, when specified by router, under
453 PP/06 Repair manualroute's ability to take options in any order, even if one
454 is the name of a transport.
457 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
459 JH/17 Change the list-building routines interface to use the expanding-string
460 triplet model, for better allocation and copying behaviour.
462 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
463 Previously it was constructed the first time a possibly-matching string
464 was met in the configuration file input during startup; now it is done
467 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
468 compatible one, to avoid the (poorly documented) possibility of a config
469 file in the working directory redirecting the DB files, possibly correpting
470 some existing file. CVE-2017-10140 assigned for BDB.
472 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
473 cache-hot. Previously, although the result was properly cached, the
474 initial verify call returned a defer.
476 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
477 the main verify for receipient in uncached-mode.
479 JH/22 Retire historical build files to an "unsupported" subdir. These are
480 defined as "ones for which we have no current evidence of testing".
482 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
483 if present. Previously it was ignored.
485 JH/24 Start using specified-initialisers in C structure init coding. This is
486 a C99 feature (it's 2017, so now considered safe).
488 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
489 if was a fixed-sized field and bitmask ops via macros; it is now more
492 PP/07 GitHub PR 56: Apply MariaDB build fix.
493 Patch provided by Jaroslav Škarvada.
495 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
496 during Coverity cleanups [4.87 JH/47]
497 Diagnosis and fix provided by Michael Fischer v. Mollard.
499 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
500 the right size to place the terminating semicolon on its own folded
501 line, the header hash was calculated to an incorrect value thanks to
502 the (relaxed) space the fold became.
504 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
505 and confused the parent.
507 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
508 which could crash as a result. This could lead to undeliverable messages.
510 JH/28 Logging: "next input sent too soon" now shows where input was truncated
513 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
514 matters on fast-turnover and PID-randomising systems, which were getting
515 out-of-order delivery.
517 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
518 a possibly-overlapping copy. The symptom was that "Remote host closed
519 connection in response to HELO" was logged instead of the actual 4xx
522 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
523 Previously only that bufferd was discarded, resulting in SYMTP command
526 JH/32 DKIM: when a message has multiple signatures matching an identity given
527 in dkim_verify_signers, run the dkim acl once for each. Previously only
528 one run was done. Bug 2189.
530 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
531 "panic the current process" to "deliberately defer". The panic log is
532 still written with the problem list name; the mail and reject logs now
533 get a temp-reject line for the message that was being handled, saying
534 something like "domains check lookup or other defer". The SMTP 451
535 message is still "Temporary local problem".
537 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
538 A crafted sequence of BDAT commands could result in in-use memory beeing
539 freed. CVE-2017-16943.
541 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
542 from SMTP input. Previously it was always done; now only done for DATA
543 and not BDAT commands. CVE-2017-16944.
545 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
546 to the message (such as an overlong header line). Previously this was
547 not done and we did not exit BDAT mode. Followon from the previous item
548 though a different problem.
554 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
555 than -2003 did; needs libidn2 in addition to libidn.
557 JH/02 The path option on a pipe transport is now expanded before use.
559 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
560 Patch provided by "Björn", documentation fix added too.
562 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
563 missing a wire-to-host endian conversion.
565 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
566 close after a BDAT command line could be taken as a following command,
567 giving a synch failure. Fix by only checking for synch immediately
568 before acknowledging the chunk.
570 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
571 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
572 macro. Patches provided by Josh Soref.
574 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
575 Previously we did not; the RFC seems ambiguous and VRFY is not listed
576 by IANA as a service extension. However, John Klensin suggests that we
579 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
580 the dkim code may be unix-mode line endings rather than smtp wire-format
581 CRLF, so prepend a CR to any bare LF.
583 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
584 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
586 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
587 the first were themselves being wrongly included in the feed into dkim
588 processing; with most chunk sizes in use this resulted in an incorrect
589 body hash calculated value.
591 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
592 DKIM signature block, for verification. Although advised against by
593 standards it is specifically not ruled illegal.
595 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
597 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
598 missing a body hash (the bh= tag).
600 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
601 It seems that HAProxy sends the Proxy Protocol information in clear and
602 only then does a TLS startup, so do the same.
604 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
605 TCP connections (such as for Spamd) unless the daemon successfully set
606 Fast Open mode on its listening sockets. This fixes breakage seen on
607 too-old kernels or those not configured for Fast Open, at the cost of
608 requiring both directions being enabled for TFO, and TFO never being used
609 by non-daemon-related Exim processes.
611 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
612 endings, at least on the first header line. Try to canonify any that get
613 past that check, despite the cost.
615 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
616 now limited to an arbitrary five deep, while parsing addresses with the
617 strip_excess_angle_brackets option enabled.
619 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
620 instead leave the unprompted TLS handshake in socket buffer for the
621 TLS library to consume.
623 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
625 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
627 JH/16 Drop variables when they go out of scope. Memory management drops a whole
628 region in one operation, for speed, and this leaves assigned pointers
629 dangling. Add checks run only under the testsuite which checks all
630 variables at a store-reset and panics on a dangling pointer; add code
631 explicitly nulling out all the variables discovered. Fixes one known
632 bug: a transport crash, where a dangling pointer for $sending_ip_address
633 originally assigned in a verify callout, is re-used.
635 PP/06 Drop '.' from @INC in various Perl scripts.
637 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
639 PP/08 Reduce a number of compilation warnings under clang; building with
640 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
641 should be warning-free.
643 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
645 HS/01 Fix portability problems introduced by PP/08 for platforms where
646 realloc(NULL) is not equivalent to malloc() [SunOS et al].
648 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
649 chunk. This allows us to accept broken chunked messages. We need a more
650 general solution here.
652 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
653 already-broken messages in the queue.
655 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
657 JH/19 Fix reference counting bug in routing-generated-address tracking.
663 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
664 supports it and a size is available (ie. the sending peer gave us one).
666 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
667 years of being deprecated). The replacements are the ACLs
668 acl_smtp_mime and acl_not_smtp_mime.
670 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
671 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
672 or even an in-clear connection were permitted. Now, if the host lookup
673 was dnssec and dane was requested then the host is only used if the
674 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
675 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
676 if one fails this test.
677 This means that a poorly-configured remote DNS will make it incommunicado;
678 but it protects against a DNS-interception attack on it.
680 JH/04 Bug 1810: make continued-use of an open smtp transport connection
681 non-noisy when a race steals the message being considered.
683 JH/05 If main configuration option tls_certificate is unset, generate a
684 self-signed certificate for inbound TLS connections.
686 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
687 in rewrites and routers.
689 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
690 and logged a warning sing 4.83; now they are a configuration file error.
692 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
693 (lacking @domain). Apply the same qualification processing as RCPT.
695 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
697 JH/10 Support ${sha256:} applied to a string (as well as the previous
700 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
701 a cutthrough deliver is pending, as we always want to make a connection.
702 This also avoids re-routing the message when later placing the cutthrough
703 connection after a verify cache hit.
704 Do not update it with the verify result either.
706 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
707 when routing results in more than one destination address.
709 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
710 signing (which inhibits the cutthrough capability). Previously only
711 the presence of an option was tested; now an expansion evaluating as
712 empty is permissible (obviously it should depend only on data available
713 when the cutthrough connection is made).
715 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
716 the relevant preceding SMTP command did not note the pipelining mode.
718 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
719 Previously they were not counted.
721 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
722 as one having no matching records. Previously we deferred the message
723 that needed the lookup.
725 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
726 distinguished as "(=".
728 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
729 for missing MX records. Previously it only worked for missing A records.
731 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
733 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
734 after the data-go-ahead and data-ack. Patch from Jason Betts.
736 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
737 even for a "none" policy. Patch from Tony Meyer.
739 JH/22 Fix continued use of a connection for further deliveries. If a port was
740 specified by a router, it must also match for the delivery to be
743 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
744 When one of the recipients of a message was unsuitable for the connection
745 (has no matching addresses), we lost track of needing to mark it
746 deferred. As a result mail would be lost.
748 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
750 JH/25 Decoding ACL controls is now done using a binary search; the source code
751 takes up less space and should be simpler to maintain. Merge the ACL
752 condition decode tables also, with similar effect.
754 JH/26 Fix problem with one_time used on a redirect router which returned the
755 parent address unchanged. A retry would see the parent address marked as
756 delivered, so not attempt the (identical) child. As a result mail would
759 JH/27 Fix a possible security hole, wherein a process operating with the Exim
760 UID can gain a root shell. Credit to http://www.halfdog.net/ for
761 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
764 JH/28 Enable {spool,log} filesystem space and inode checks as default.
765 Main config options check_{log,spool}_{inodes,space} are now
766 100 inodes, 10MB unless set otherwise in the configuration.
768 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
769 Previously it only applied to the main-section connection policy
772 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
774 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
775 by me. Added RFC7919 DH primes as an alternative.
777 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
778 are not in the system include path.
780 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
781 GnuTLS, when a session startup failed (eg because the client disconnected)
782 Exim did stdio operations after fclose. This was exposed by a recent
783 change which nulled out the file handle after the fclose.
785 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
786 signed directly by the cert-signing cert, rather than an intermediate
787 OCSP-signing cert. This is the model used by LetsEncrypt.
789 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
791 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
792 an incoming connection.
794 HS/02 Bug 1802: Do not half-close the connection after sending a request
797 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
798 fallback to "prime256v1".
800 JH/34 SECURITY: Use proper copy of DATA command in error message.
801 Could leak key material. Remotely exploitable. CVE-2016-9963.
807 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
808 and 3.4.4 - once the server is enabled to respond to an OCSP request
809 it does even when not requested, resulting in a stapling non-aware
810 client dropping the TLS connection.
812 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
813 support variable-length bit vectors. No functional change.
815 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
816 The I= interface field on outgoing lines is now after the H= remote
817 host field, same as incoming lines. There is a separate
818 outgoing_interface log selector which allows you to disable the
821 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
822 If not running log_selector +smtp_connection the mainlog would be held
823 open indefinitely after a "too many connections" event, including to a
824 deleted file after a log rotate. Leave the per net connection logging
825 leaving it open for efficiency as that will be quickly detected by the
826 check on the next write.
828 HS/01 Bug 1671: Fix post transport crash.
829 Processing the wait-<transport> messages could crash the delivery
830 process if the message IDs didn't exist for some reason. When
831 using 'split_spool_directory=yes' the construction of the spool
832 file name failed already, exposing the same netto behaviour.
834 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
835 mime_regex ACL conditions.
837 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
838 to DSN fail messages (bounces): remote IP, remote greeting, remote response
839 to HELO, local diagnostic string.
841 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
842 log line to debug. Even when configured with a tls authenticator many
843 client connections are expected to not authenticate in this way, so
844 an authenticate fail is not an error.
846 HS/02 Add the Exim version string to the process info. This way exiwhat
847 gives some more detail about the running daemon.
849 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
850 matter for fast-change records such as DNSBLs.
852 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
853 retry record, even if constant. There may be multiple transports with
854 different interface settings and the retry behaviour needs to be kept
857 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
859 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
861 JH/10 Bug 840: fix log_defer_output option of pipe transport
863 JH/11 Bug 830: use same host for all RCPTS of a message, even under
864 hosts_randomize. This matters a lot when combined with mua_wrapper.
866 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
867 ${quote_pgsql:<string>} operator.
869 JH/13 Bug 1708: avoid misaligned access in cached lookup.
871 JH/14 Change header file name for freeradius-client. Relevant if compiling
872 with Radius support; from the Gentoo tree and checked under Fedora.
874 JH/15 Bug 1712: Introduce $prdr_requested flag variable
876 JH/16 Bug 1714: Permit an empty string as expansion result for transport
877 option transport_filter, meaning no filtering.
879 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
881 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
882 defaults to "*" (all hosts). The variable is now available when not built
883 with TLS, default unset, mainly to enable keeping the testsuite sane.
884 If a server certificate is not supplied (via tls_certificate) an error is
885 logged, and clients will find TLS connections fail on startup. Presumably
886 they will retry in-clear.
887 Packagers of Exim are strongly encouraged to create a server certificate
888 at installation time.
890 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
891 with the $config_file variable.
893 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
894 in transport context, after the attempt, and per-recipient. The latter type
895 is per host attempted. The event data is the error message, and the errno
896 information encodes the lookup type (A vs. MX) used for the (first) host,
897 and the trailing two digits of the smtp 4xx response.
899 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
900 to write to mainlog (or rejectlog, paniclog) in the window between file
901 creation and permissions/ownership being changed. Particularly affects
902 installations where exicyclog is run as root, rather than exim user;
903 result is that the running daemon panics and dies.
905 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
907 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
908 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
909 "pri" and "weight". Note that the previous implicit priority given by the
910 list order is no longer honoured.
912 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
915 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
916 by defining SUPPORT_SOCKS.
918 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
919 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
920 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
921 variables are renamed to proxy_{local,external}_{address,port}.
923 JH/27 Move Internationalisation support from Experimental to mainline, enabled
924 for a build by defining SUPPORT_I18N
926 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
927 of the query string, and make ${quote_redis:} do that quoting.
929 JH/29 Move Events support from Experimental to mainline, enabled by default
930 and removable for a build by defining DISABLE_EVENT.
932 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
934 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
935 cached by the daemon.
937 JH/32 Move Redis support from Experimental to mainline, enabled for a build
938 by defining LOOKUP_REDIS. The libhiredis library is required.
940 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
941 keys are given for lookup.
943 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
944 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
945 only supported when built with TLS support. The PolarSSL SHA routines
946 are still used when the TLS library is too old for convenient support.
948 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
949 openssl_options), for security. OpenSSL forces this from version 1.1.0
950 server-side so match that on older versions.
952 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
953 allocation for $value could be released as the expansion processing
954 concluded, but leaving the global pointer active for it.
956 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
957 and to use the domains and local_parts ACL conditions.
959 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
960 incorrectly not doubled on cutthrough transmission, hence seen as a
961 body-termination at the receiving system - resulting in truncated mails.
962 Commonly the sender saw a TCP-level error, and retransmitted the message
963 via the normal store-and-forward channel. This could result in duplicates
964 received - but deduplicating mailstores were liable to retain only the
965 initial truncated version.
967 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
969 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
971 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
972 we're in there, support oversigning also; bug 1309.
974 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
976 HS/04 Add support for keep_environment and add_environment options.
978 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
979 either intentional arithmetic overflow during PRNG, or testing config-
982 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
983 delivery resulted in actual delivery. Cancel cutthrough before DATA
986 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
987 rejects a recipient: pass the reject to the originator.
989 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
990 Many were false-positives and ignorable, but it's worth fixing the
993 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
994 for the new environment-manipulation done at startup. Move the routines
995 from being local to tls.c to being global via the os.c file.
997 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
998 an extract embedded as result-arg for a map, the first arg for extract
999 is unavailable so we cannot tell if this is a numbered or keyed
1000 extraction. Accept either.
1006 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1009 JH/02 The smtp transport option "multi_domain" is now expanded.
1011 JH/03 The smtp transport now requests PRDR by default, if the server offers
1014 JH/04 Certificate name checking on server certificates, when exim is a client,
1015 is now done by default. The transport option tls_verify_cert_hostnames
1016 can be used to disable this per-host. The build option
1017 EXPERIMENTAL_CERTNAMES is withdrawn.
1019 JH/05 The value of the tls_verify_certificates smtp transport and main options
1020 default to the word "system" to access the system default CA bundle.
1021 For GnuTLS, only version 3.0.20 or later.
1023 JH/06 Verification of the server certificate for a TLS connection is now tried
1024 (but not required) by default. The verification status is now logged by
1025 default, for both outbound TLS and client-certificate supplying inbound
1028 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1031 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1032 Status Notification (bounce) messages are now MIME format per RFC 3464.
1033 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1034 under the control of the dsn_advertise_hosts option, and routers may
1035 have a dsn_lasthop option.
1037 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1038 default, modifiable by a malware= option. The list separator for
1039 the options can now be changed in the usual way. Bug 68.
1041 JH/10 The smtp_receive_timeout main option is now expanded before use.
1043 JH/11 The incoming_interface log option now also enables logging of the
1044 local interface on delivery outgoing connections.
1046 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1047 if the interface and destination host and port all match.
1049 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1052 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1053 Patch from Andrew Lewis.
1055 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1056 now supports optional time-restrictions, weighting, and priority
1057 modifiers per server. Patch originally by <rommer@active.by>.
1059 JH/16 The spamd_address main option now supports a mixed list of local
1060 and remote servers. Remote servers can be IPv6 addresses, and
1061 specify a port-range.
1063 JH/17 Bug 68: The spamd_address main option now supports an optional
1064 timeout value per server.
1066 JH/18 Bug 1581: Router and transport options headers_add/remove can
1067 now have the list separator specified.
1069 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1072 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1075 JH/21 Support for the A6 type of dns record is withdrawn.
1077 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1078 rather than the verbs used.
1080 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1081 from 255 to 1024 chars.
1083 JH/24 Verification callouts now attempt to use TLS by default.
1085 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1086 are generic router options now. The defaults didn't change.
1088 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1089 Original patch from Alexander Shikoff, worked over by JH.
1091 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1092 files if ZCAT_COMMAND is not executable.
1094 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1096 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1098 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1099 Normally benign, it bites when the pair was led to by a CNAME;
1100 modern usage is to not canonicalize the domain to a CNAME target
1101 (and we were inconsistent anyway for A-only vs AAAA+A).
1103 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1105 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1106 when evaluating $sender_host_dnssec.
1108 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1109 $sender_helo_dnssec variable.
1111 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1113 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1115 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1117 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1118 documented as working, but never had. Support all but $spam_report.
1120 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1121 added for tls authenticator.
1123 HS/03 Add perl_taintmode main config option
1129 TL/01 When running the test suite, the README says that variables such as
1130 no_msglog_check are global and can be placed anywhere in a specific
1131 test's script, however it was observed that placement needed to be near
1132 the beginning for it to behave that way. Changed the runtest perl
1133 script to read through the entire script once to detect and set these
1134 variables, reset to the beginning of the script, and then run through
1135 the script parsing/test process like normal.
1137 TL/02 The BSD's have an arc4random API. One of the functions to induce
1138 adding randomness was arc4random_stir(), but it has been removed in
1139 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1140 function when detected.
1142 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1143 cause callback expansion.
1145 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1146 syntax errors in an expansion can be treated as a string instead of
1147 logging or causing an error, due to the internal use of bool_lax
1148 instead of bool when processing it.
1150 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1151 server certificates when making smtp deliveries.
1153 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1155 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1157 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1159 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1160 Merged patch from Sebastian Wiedenroth.
1162 JH/05 Fix results-pipe from transport process. Several recipients, combined
1163 with certificate use, exposed issues where response data items split
1164 over buffer boundaries were not parsed properly. This eventually
1165 resulted in duplicates being sent. This issue only became common enough
1166 to notice due to the introduction of connection certificate information,
1167 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1169 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1170 size buffer was used, resulting in syntax errors when an expansion
1173 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1174 version 3.3.6 or later.
1176 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1177 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1178 both become "event_action", the variables become $event_name, $event_data
1179 and $event_defer_errno. There is a new variable $verify_mode, usable in
1180 routers, transports and related events. The tls:cert event is now also
1181 raised for inbound connections, if the main configuration event_action
1184 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1185 early OCSP support, but no stapling (appears to be less than 1.0.0).
1187 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1188 server certificate names available under the smtp transport option
1189 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1192 JH/10 Time-related extraction expansions from certificates now use the main
1193 option "timezone" setting for output formatting, and are consistent
1194 between OpenSSL and GnuTLS compilations. Bug 1541.
1196 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1197 encoded parameter in the incoming message. Bug 1558.
1199 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1200 include certificate info, eximon was claiming there were spoolfile
1203 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1205 JH/14 Log delivery-related information more consistently, using the sequence
1206 "H=<name> [<ip>]" wherever possible.
1208 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1209 are problematic for Debian distribution, omit them from the release
1212 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1214 JH/16 Fix string representation of time values on 64bit time_t architectures.
1217 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1218 output list separator was used.
1223 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1224 checkers that were complaining about end of non-void function with no
1227 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1228 This was a regression introduced in 4.83 by another bugfix.
1230 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1232 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1233 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1239 TF/01 Correctly close the server side of TLS when forking for delivery.
1241 When a message was received over SMTP with TLS, Exim failed to clear up
1242 the incoming connection properly after forking off the child process to
1243 deliver the message. In some situations the subsequent outgoing
1244 delivery connection happened to have the same fd number as the incoming
1245 connection previously had. Exim would try to use TLS and fail, logging
1246 a "Bad file descriptor" error.
1248 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1249 utilities have not been installed.
1251 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1252 temporary space as the ACL may create new global variables.
1254 TL/01 LDAP support uses per connection or global context settings, depending
1255 upon the detected version of the libraries at build time.
1257 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1258 to extract and use the src ip:port in logging and expansions as if it
1259 were a direct connection from the outside internet. PPv2 support was
1260 updated based on HAProxy spec change in May 2014.
1262 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1264 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1265 Properly escape header and check for NULL return.
1267 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1270 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1272 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1273 characters in header names, implemented as a verify condition.
1274 Contributed by Michael Fischer v. Mollard.
1276 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1277 results permerror and temperror. Previous values are deprecated but
1278 still accepted. In a future release, err_perm and err_temp will be
1279 completely removed, which will be a backward incompatibility if the
1280 ACL tests for either of these two old results. Patch contributed by
1281 user bes-internal on the mailing list.
1283 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1285 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1286 selectors, in both main and reject logs.
1288 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1291 JH/07 Add malware type "sock" for talking to simple daemon.
1293 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1295 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1296 routers/transports under cutthrough routing.
1298 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1299 numbers. Touch up "bool" conditional to keep the same definition.
1301 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1303 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1305 JH/12 Expand items in router/transport headers_add or headers_remove lists
1306 individually rather than the list as a whole. Bug 1452.
1308 Required for reasonable handling of multiple headers_ options when
1309 they may be empty; requires that headers_remove items with embedded
1310 colons must have them doubled (or the list-separator changed).
1312 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1313 view the policy declared in the DMARC record. Currently, $dmarc_status
1314 is a combined value of both the record presence and the result of the
1317 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1319 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1320 dnslookup router and the smtp transport (applying to the forward
1323 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1324 of ldap servers used for a specific lookup. Patch provided by Heiko
1327 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1328 New variable $lookup_dnssec_authenticated for observability.
1330 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1331 Patch submitted by Lars Timman.
1333 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1335 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1336 Requires trusted mode and valid format message id, aborts otherwise.
1337 Patch contributed by Heiko Schlichting.
1339 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1340 certextract with support for various fields. Bug 1358.
1342 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1343 is requested by default, modifiable by smtp transport option
1346 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1347 operate on certificate variables to give certificate fingerprints
1348 Also new ${sha256:cert_variable}.
1350 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1352 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1353 Christian Aistleitner.
1355 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1357 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1358 file. Patch from Wolfgang Breyha.
1360 JH/25 Expand the coverage of the delivery $host and $host_address to
1361 client authenticators run in verify callout. Bug 1476.
1363 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1364 align with daemon_smtp_ports. Bug 72.
1366 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1367 support and error reporting did not work properly.
1369 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1370 and is readable. Patch from Andrew Colin Kissa.
1372 TL/14 Enhance documentation of ${run expansion and how it parses the
1373 commandline after expansion, particularly in the case when an
1374 unquoted variable expansion results in an empty value.
1376 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1378 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1379 by renaming away from it. Addresses GH issue 15, reported by
1382 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1384 TL/15 SECURITY: prevent double expansion in math comparison functions
1385 (can expand unsanitized data). Not remotely exploitable.
1392 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1394 PP/02 Make -n do something, by making it not do something.
1395 When combined with -bP, the name of an option is not output.
1397 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1400 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1401 $sender_host_name and config options to manage this, and basic check
1404 PP/05 DSCP support for outbound connections and control modifier for inbound.
1406 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1407 (Only plugin which currently uses this is kerberos4, which nobody should
1408 be using, but we should make it available and other future plugins might
1409 conceivably use it, even though it would break NAT; stuff *should* be
1410 using channel bindings instead).
1412 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1413 name; added for Sendmail compatibility; requires admin caller.
1414 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1415 just ignore it); requires trusted caller.
1416 Also parse but ignore: -Ac -Am -X<logfile>
1419 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1421 TL/02 Add +smtp_confirmation as a default logging option.
1423 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1424 Patch by Magnus Holmgren from 2007-02-20.
1426 TL/04 Bugzilla 1281 - Spec typo.
1427 Bugzilla 1283 - Spec typo.
1428 Bugzilla 1290 - Spec grammar fixes.
1430 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1432 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1434 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1435 mailing list by by Dmitry Isaikin.
1437 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1439 JH/02 Support "G" suffix to numbers in ${if comparisons.
1441 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1443 NM/01 Bugzilla 1197 - Spec typo
1444 Bugzilla 1196 - Spec examples corrections
1446 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1448 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1449 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1452 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1453 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1455 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1456 "acl {{name}{arg}...}", and optional args on acl condition
1459 JH/05 Permit multiple router/transport headers_add/remove lines.
1461 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1463 JH/07 Avoid using a waiting database for a single-message-only transport.
1464 Performance patch from Paul Fisher. Bugzilla 1262.
1466 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1469 JH/09 Add $headers_added variable, with content from use of ACL modifier
1470 add_header (but not yet added to the message). Bugzilla 199.
1472 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1473 Pulled from Bugzilla 817 by Wolfgang Breyha.
1475 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1477 (nb: this is the same fix as in Exim 4.80.1)
1479 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1482 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1485 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1487 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1488 advertise SMTP AUTH mechanism to us, instead of a generic
1489 protocol violation error. Also, make Exim more robust to bad
1490 data from the Dovecot auth socket.
1492 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1494 When a queue runner is handling a message, Exim first routes the
1495 recipient addresses, during which it prunes them based on the retry
1496 hints database. After that it attempts to deliver the message to
1497 any remaining recipients. It then updates the hints database using
1500 So if a recipient address works intermittently, it can get repeatedly
1501 deferred at routing time. The retry hints record remains fresh so the
1502 address never reaches the final cutoff time.
1504 This is a fairly common occurrence when a user is bumping up against
1505 their storage quota. Exim had some logic in its local delivery code
1506 to deal with this. However it did not apply to per-recipient defers
1507 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1509 This change adds a proper retry rule check during routing so that the
1510 final cutoff time is checked against the message's age. We only do
1511 this check if there is an address retry record and there is not a
1512 domain retry record; this implies that previous attempts to handle
1513 the address had the retry_use_local_parts option turned on. We use
1514 this as an approximation for the destination being like a local
1515 delivery, as in LMTP.
1517 I suspect this new check makes the old local delivery cutoff check
1518 redundant, but I have not verified this so I left the code in place.
1520 TF/02 Correct gecos expansion when From: is a prefix of the username.
1522 Test 0254 submits a message to Exim with the header
1526 When I ran the test suite under the user fanf2, Exim expanded
1527 the header to contain my full name, whereas it should have added
1528 a Resent-Sender: header. It erroneously treats any prefix of the
1529 username as equal to the username.
1531 This change corrects that bug.
1533 GF/01 DCC debug and logging tidyup
1534 Error conditions log to paniclog rather than rejectlog.
1535 Debug lines prefixed by "DCC: " to remove any ambiguity.
1537 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1539 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1540 Bug spotted by Jeremy Harris; was flawed since initial commit.
1541 Would have resulted in OCSP responses post-SNI triggering an Exim
1542 NULL dereference and crash.
1544 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1546 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1547 Bug detection, analysis and fix by Samuel Thibault.
1548 Bugzilla 1331, Debian bug #698092.
1550 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1552 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1553 Server implementation by Todd Lyons, client by JH.
1554 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1555 config variable "prdr_enable" controls whether the server
1556 advertises the facility. If the client requests PRDR a new
1557 acl_data_smtp_prdr ACL is called once for each recipient, after
1558 the body content is received and before the acl_smtp_data ACL.
1559 The client is controlled by both of: a hosts_try_prdr option
1560 on the smtp transport, and the server advertisement.
1561 Default client logging of deliveries and rejections involving
1562 PRDR are flagged with the string "PRDR".
1564 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1565 fclose(). Diagnosis by Todd Lyons.
1567 PP/17 Update configure.default to handle IPv6 localhost better.
1568 Patch by Alain Williams (plus minor tweaks).
1571 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1572 This is now consistent with GnuTLS, and is now documented: the
1573 previous undocumented portable approach to treating the option as
1574 unset was to force an expansion failure. That still works, and
1575 an empty string is now equivalent.
1577 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1578 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1579 not performing validation itself.
1581 PP/20 Added force_command boolean option to pipe transport.
1582 Patch from Nick Koston, of cPanel Inc.
1584 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1587 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1589 PP/21 Fix eximon continuous updating with timestamped log-files.
1590 Broken in a format-string cleanup in 4.80, missed when I repaired the
1591 other false fix of the same issue.
1592 Report and fix from Heiko Schlichting.
1595 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1596 Report from Prashanth Katuri.
1598 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1599 It's SecureTransport, so affects any MacOS clients which use the
1600 system-integrated TLS libraries, including email clients.
1602 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1603 using a MIME ACL for non-SMTP local injection.
1604 Report and assistance in diagnosis by Warren Baker.
1606 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1608 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1610 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1611 last id that failed so it may be referenced in subsequent ACL's.
1613 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1616 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1617 ldap library initialization, allowing self-signed CA's to be
1618 used. Also properly sets require_cert option later in code by
1619 using NULL (global ldap config) instead of ldap handle (per
1620 session). Bug diagnosis and testing by alxgomz.
1622 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1623 the src/util/ subdirectory.
1625 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1626 renamed to Transport Post Delivery Action by Jeremy Harris, as
1629 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1630 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1631 redis_servers = needs to be configured which will be used by the redis
1632 lookup. Patch from Warren Baker, of The Packet Hub.
1634 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1636 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1637 hostname or reverse DNS when processing a host list. Used suggestions
1638 from multiple comments on this bug.
1640 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1642 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1643 Missed a few lines, added it to make the runtest require no keyboard
1646 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1647 contains upper case chars. Make router use caseful_local_part.
1649 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1650 support when GnuTLS has been built with p11-kit.
1656 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1658 This, or similar/improved, will also be change PP/11 of 4.82.
1664 PP/01 Handle short writes when writing local log-files.
1665 In practice, only affects FreeBSD (8 onwards).
1666 Bugzilla 1053, with thanks to Dmitry Isaikin.
1668 NM/01 Bugzilla 949 - Documentation tweak
1670 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1673 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1675 PP/02 Implemented gsasl authenticator.
1677 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1679 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1680 `pkg-config foo` for cflags/libs.
1682 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1683 with rest of GSASL and with heimdal_gssapi.
1685 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1686 `pkg-config foo` for cflags/libs for the TLS implementation.
1688 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1689 properties get this fed in as external SSF. A number of robustness
1690 and debugging improvements to the cyrus_sasl authenticator.
1692 PP/08 cyrus_sasl server now expands the server_realm option.
1694 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1695 Patch by Jeremy Harris.
1697 PP/10 Added dbmjz lookup type.
1699 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1701 PP/12 MAIL args handles TAB as well as SP, for better interop with
1702 non-compliant senders.
1703 Analysis and variant patch by Todd Lyons.
1705 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1706 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1707 Patch from Dirk Mueller <dmueller@suse.com>
1709 PP/13 tls_peerdn now print-escaped for spool files.
1710 Observed some $tls_peerdn in wild which contained \n, which resulted
1711 in spool file corruption.
1713 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1714 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1715 or write after TLS renegotiation, which otherwise led to messages
1718 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1719 as a tracking header (ie: a signed header comes before the signature).
1720 Patch from Wolfgang Breyha.
1722 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1723 comma-sep list; embedded commas doubled.
1725 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1727 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1729 Report and patch from Dmitry Banschikov.
1731 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1732 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1733 failures. We appear to now support TLS1.1+ with Exim.
1735 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1736 lets Exim select keys and certificates based upon TLS SNI from client.
1737 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1738 before an outbound SMTP session. New log_selector, +tls_sni.
1740 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1741 NULL dereference. Report and patch from Alun Jones.
1743 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1744 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1745 Not seeing resolver debug output on NetBSD, but suspect this is a
1746 resolver implementation change.
1748 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1749 Left warnings. Added "eximon gdb" invocation mode.
1751 PP/21 Defaulting "accept_8bitmime" to true, not false.
1753 PP/22 Added -bw for inetd wait mode support.
1755 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1756 locate the relevant includes and libraries. Made this the default.
1758 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1759 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1761 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1762 This may cause build issues on older platforms.
1764 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1765 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1766 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1767 Added SNI support via GnuTLS too.
1768 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1770 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1772 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1773 Applied second patch from Janne, implementing suggestion to default
1774 multiple-strings-in-record handling to match SPF spec.
1776 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1778 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1779 read-only, out of scope).
1780 Patch from Wolfgang Breyha, report from Stuart Northfield.
1782 PP/29 Fix three issues highlighted by clang analyser static analysis.
1783 Only crash-plausible issue would require the Cambridge-specific
1784 iplookup router and a misconfiguration.
1785 Report from Marcin Mirosław.
1787 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1789 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1790 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1791 As part of this, removing so much warning spew let me fix some minor
1792 real issues in debug logging.
1794 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1795 assignment on my part. Fixed.
1797 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1798 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1799 Janne Snabb (who went above and beyond: thank you).
1801 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1802 string otherwise requires a connection and a bunch more work and it's
1803 relatively easy to get wrong. Should also expose TLS library linkage
1806 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1807 64-bit ${eval} (JH/03).
1809 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1810 GNU libc to support some of the 64-bit stuff, should not lead to
1811 conflicts. Defined before os.h is pulled in, so if a given platform
1812 needs to override this, it can.
1814 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1815 protection layer was required, which is not implemented.
1816 Bugzilla 1254, patch from Wolfgang Breyha.
1818 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1819 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1820 tls_dhparam take prime identifiers. Also unbreak combination of
1821 OpenSSL+DH_params+TLSSNI.
1823 PP/39 Disable SSLv2 by default in OpenSSL support.
1829 PP/01 Solaris build fix for Oracle's LDAP libraries.
1830 Bugzilla 1109, patch from Stephen Usher.
1832 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1834 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1837 TF/02 Fix a couple more cases where we did not log the error message
1838 when unlink() failed. See also change 4.74-TF/03.
1840 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1841 lock up or crash if it happened to be inside a call to libc when it
1842 got a SIGUSR1 from exiwhat.
1844 The SIGUSR1 handler appends the current process status to the process
1845 log which is later printed by exiwhat. It used to use the general
1846 purpose logging code to do this, but several functions it calls are
1847 not safe for signals.
1849 The new output code in the SIGUSR1 handler is specific to the process
1850 log, and simple enough that it's easy to inspect for signal safety.
1851 Removing some special cases also simplifies the general logging code.
1852 Removing the spurious timestamps from the process log simplifies
1855 TF/04 Improved ratelimit ACL condition.
1857 The /noupdate option has been deprecated in favour of /readonly which
1858 has clearer semantics. The /leaky, /strict, and /readonly update modes
1859 are mutually exclusive. The update mode is no longer included in the
1860 database key; it just determines when the database is updated. (This
1861 means that when you upgrade Exim will forget old rate measurements.)
1863 Exim now checks that the per_* options are used with an update mode that
1864 makes sense for the current ACL. For example, when Exim is processing a
1865 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1866 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1867 must specify per_mail/readonly. If you omit the update mode it defaults to
1868 /leaky where that makes sense (as before) or /readonly where required.
1870 The /noupdate option is now undocumented but still supported for
1871 backwards compatibility. It is equivalent to /readonly except that in
1872 ACLs where /readonly is required you may specify /leaky/noupdate or
1873 /strict/noupdate which are treated the same as /readonly.
1875 A useful new feature is the /count= option. This is a generalization
1876 of the per_byte option, so that you can measure the throughput of other
1877 aggregate values. For example, the per_byte option is now equivalent
1878 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1880 The per_rcpt option has been generalized using the /count= mechanism
1881 (though it's more complicated than the per_byte equivalence). When it is
1882 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1883 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1884 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1885 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1886 Note that using per_rcpt with a non-readonly update mode in more than
1887 one ACL will cause the recipients to be double-counted. (The per_mail
1888 and per_byte options don't have this problem.)
1890 The handling of very low rates has changed slightly. If the computed rate
1891 is less than the event's count (usually one) then this event is the first
1892 after a long gap. In this case the rate is set to the same as this event's
1893 count, so that the first message of a spam run is counted properly.
1895 The major new feature is a mechanism for counting the rate of unique
1896 events. The new per_addr option counts the number of different
1897 recipients that someone has sent messages to in the last time period. It
1898 behaves like per_rcpt if all the recipient addresses are different, but
1899 duplicate recipient addresses do not increase the measured rate. Like
1900 the /count= option this is a general mechanism, so the per_addr option
1901 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1902 example, measure the rate that a client uses different sender addresses
1903 with the options per_mail/unique=$sender_address. There are further
1904 details in the main documentation.
1906 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1908 TF/06 Removed a few PCRE remnants.
1910 TF/07 Automatically extract Exim's version number from tags in the git
1911 repository when doing development or release builds.
1913 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1914 Bugzilla 879. Patch from Paul Fisher.
1916 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1917 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1920 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1922 PP/05 Variable $av_failed, true if the AV scanner deferred.
1923 Bugzilla 1078. Patch from John Horne.
1925 PP/06 Stop make process more reliably on build failure.
1926 Bugzilla 1087. Patch from Heiko Schlittermann.
1928 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1929 Bugzilla 1089. Patch from Heiko Schlittermann.
1931 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1932 Bugzilla 1131. Patch from Holger Weiß.
1934 PP/09 Handle IPv6 addresses with SPF.
1935 Bugzilla 860. Patch from Wolfgang Breyha.
1937 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1939 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1942 PP/11 match_* no longer expand right-hand-side by default.
1943 New compile-time build option, EXPAND_LISTMATCH_RHS.
1944 New expansion conditions, "inlist", "inlisti".
1946 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1948 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1950 PP/14 fix log_write() format string regression from TF/03.
1951 Bugzilla 1152. Patch from Dmitry Isaikin.
1957 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1959 PP/02 Harmonised TLS library version reporting; only show if debugging.
1960 Layout now matches that introduced for other libraries in 4.74 PP/03.
1962 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1964 PP/04 New "dns_use_edns0" global option.
1966 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1969 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1970 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1972 TK/01 Updated PolarSSL code to 0.14.2.
1973 Bugzilla 1097. Patch from Andreas Metzler.
1975 PP/07 Catch divide-by-zero in ${eval:...}.
1976 Fixes bugzilla 1102.
1978 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1981 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1982 format-string attack -- SECURITY: remote arbitrary code execution.
1984 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1985 time unintentionally subject to list matching rules, letting the header
1986 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1987 arbitrary string expansion). This allowed for information disclosure.
1989 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1990 INT_MIN/-1 -- value coerced to INT_MAX.
1996 NM/01 Workaround for PCRE version dependency in version reporting
1999 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2000 This fixes portability to compilers other than gcc, notably
2001 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2003 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2004 makefiles for portability to HP-UX and POSIX correctness.
2006 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2007 Also via indented variable definition in the Makefile.
2008 (Debugging by Oliver Heesakkers).
2010 PP/02 Restore caching of spamd results with expanded spamd_address.
2011 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2013 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2014 Improves build reliability. Fix from: Frank Elsner
2016 NM/02 Fix wide character breakage in the rfc2047 coding
2017 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2019 NM/03 Allow underscore in dnslist lookups
2020 Fixes bug 1026. Patch from Graeme Fowler
2022 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2023 Code patches from Adam Ciarcinski of NetBSD.
2025 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2028 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2029 is logged, to avoid truncation. Patch from John Horne.
2031 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2032 Patch from Jakob Hirsch.
2034 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2035 SQL string expansion failure details.
2036 Patch from Andrey Oktyabrski.
2038 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2039 Patch from Simon Arlott.
2041 PP/09 New lookups functionality failed to compile on old gcc which rejects
2042 extern declarations in function scope.
2043 Patch from Oliver Fleischmann
2045 PP/10 Use sig_atomic_t for flags set from signal handlers.
2046 Check getgroups() return and improve debugging.
2047 Fixed developed for diagnosis in bug 927 (which turned out to be
2050 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2051 Patch from Mark Zealey.
2053 PP/12 Bugzilla 1056: Improved spamd server selection.
2054 Patch from Mark Zealey.
2056 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2057 Based on patch from Heiko Schlittermann.
2059 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2060 Patch from Uwe Doering, sign-off by Michael Haardt.
2062 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2063 variable declaration deep within a block. Bug and patch from
2066 PP/15 lookups-Makefile IRIX compatibility coercion.
2068 PP/16 Make DISABLE_DKIM build knob functional.
2070 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2071 Patch by Simon Arlott
2073 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2074 variable argument macros. Our copy now differs from upstream.
2080 TF/01 Failure to get a lock on a hints database can have serious
2081 consequences so log it to the panic log.
2083 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2084 controlled using the smtp_confirmation log selector.
2086 TF/03 Include the error message when we fail to unlink a spool file.
2088 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2089 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2090 for maintaining out-of-tree patches for some time.
2092 PP/01 Bugzilla 139: Documentation and portability issues.
2093 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2094 Handle per-OS dynamic-module compilation flags.
2096 PP/02 Let /dev/null have normal permissions.
2097 The 4.73 fixes were a little too stringent and complained about the
2098 permissions on /dev/null. Exempt it from some checks.
2099 Reported by Andreas M. Kirchwitz.
2101 PP/03 Report version information for many libraries, including
2102 Exim version information for dynamically loaded libraries. Created
2103 version.h, now support a version extension string for distributors
2104 who patch heavily. Dynamic module ABI change.
2106 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2107 privilege escalation vulnerability whereby the Exim run-time user
2108 can cause root to append content of the attacker's choosing to
2111 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2114 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2115 If dropping privileges for untrusted macros, we disabled normal logging
2116 on the basis that it would fail; for the Exim run-time user, this is not
2117 the case, and it resulted in successful deliveries going unlogged.
2118 Fixed. Reported by Andreas Metzler.
2124 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2125 only prepend for the Resent-* case. Fixes regression introduced in
2126 Exim 4.70 by NM/22 for Bugzilla 607.
2128 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2129 increasing numbers of administrators be bitten by this.
2131 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2133 PP/03 Bugzilla 994: added openssl_options main configuration option.
2135 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2137 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2139 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2141 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2142 without a peer certificate, leading to a segfault because of an
2143 assumption that peers always have certificates. Be a little more
2144 paranoid. Problem reported by Martin Tscholak.
2146 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2147 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2148 NB: ClamAV planning to remove STREAM in "middle of 2010".
2149 CL also introduces -bmalware, various -d+acl logging additions and
2150 more caution in buffer sizes.
2152 PP/09 Implemented reverse_ip expansion operator.
2154 PP/10 Bugzilla 937: provide a "debug" ACL control.
2156 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2158 PP/12 Bugzilla 973: Implement --version.
2160 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2162 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2164 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2166 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2167 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2168 ignore trailing whitespace.
2170 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2172 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2175 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2176 Notification from Dr Andrew Aitchison.
2178 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2179 ExtendedDetectionInfo response format.
2180 Notification from John Horne.
2182 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2185 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2186 XSL and documented dependency on system catalogs, with examples of how
2189 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2192 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2193 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2194 configuration file which is writeable by the Exim user or group.
2196 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2197 of configuration files to cover files specified with the -C option if
2198 they are going to be used with root privileges, not just the default
2201 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2202 option (effectively making it always true).
2204 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2205 files to be used while preserving root privileges.
2207 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2208 that rogue child processes cannot use them.
2210 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2211 run-time user, instead of root.
2213 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2214 Exim run-time user without dropping privileges.
2216 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2217 result string, instead of calling string_vformat() twice with the same
2220 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2221 for other users. Others should always drop root privileges if they use
2222 -C on the command line, even for a whitelisted configure file.
2224 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2226 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2232 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2233 $data_path, and $header_path variables; fixed documentation bugs and
2236 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2237 exipick to access non-standard spools, including the "frozen" queue
2240 NM/01 Bugzilla 965: Support mysql stored procedures.
2241 Patch from Alain Williams
2243 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2245 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2246 Patch from Andreas Metzler
2248 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2249 Patch from Kirill Miazine
2251 NM/05 Bugzilla 671: Added umask to procmail example.
2253 JJ/03 installed exipick 20100323.0, fixing doc bug
2255 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2256 directory. Notification and patch from Dan Rosenberg.
2258 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2260 TK/02 Improve log output when DKIM signing operation fails.
2262 MH/01 Treat the transport option dkim_domain as a colon separated
2263 list, not as a single string, and sign the message with each element,
2264 omitting multiple occurences of the same signer.
2266 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2267 Bugzilla 985, 986. Patch by Simon Arlott
2269 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2270 Patch by Simon Arlott
2272 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2273 MBX locking. Notification from Dan Rosenberg.
2279 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2281 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2283 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2285 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2287 NM/04 Bugzilla 915: Flag broken perl installation during build.
2293 TK/01 Added patch by Johannes Berg that expands the main option
2294 "spamd_address" if it starts with a dollar sign.
2296 TK/02 Write list of recipients to X-Envelope-Sender header when building
2297 the mbox-format spool file for content scanning (suggested by Jakob
2300 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2301 (http://www.dcc-servers.net/) support via dccifd. Activated by
2302 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2304 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2305 by Mark Daniel Reidel <mr@df.eu>.
2307 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2308 When building exim an external PCRE library is now needed -
2309 PCRE is a system library on the majority of modern systems.
2310 See entry on PCRE_LIBS in EDITME file.
2312 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2313 conversation. Added nologin parameter to request.
2314 Patch contributed by Kirill Miazine.
2316 TF/01 Do not log submission mode rewrites if they do not change the address.
2318 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2320 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2321 log files in place. Contributed by Roberto Lima.
2323 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2325 TF/03 Bugzilla 615: When checking the local_parts router precondition
2326 after a local_part_suffix or local_part_prefix option, Exim now
2327 does not use the address's named list lookup cache, since this
2328 contains cached lookups for the whole local part.
2330 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2331 Robert Millan. Documentation is in experimental-spec.txt.
2333 TF/04 Bugzilla 668: Fix parallel build (make -j).
2335 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2337 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2338 Patch provided by Jan Srzednicki.
2340 TF/05 Leading white space used to be stripped from $spam_report which
2341 wrecked the formatting. Now it is preserved.
2343 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2344 that they are available at delivery time.
2346 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2348 TF/08 TLS error reporting now respects the incoming_interface and
2349 incoming_port log selectors.
2351 TF/09 Produce a more useful error message if an SMTP transport's hosts
2352 setting expands to an empty string.
2354 NM/06 Bugzilla 744: EXPN did not work under TLS.
2355 Patch provided by Phil Pennock.
2357 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2358 Patch provided by Richard Godbee.
2360 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2361 acl_smtp_notquit, added index entry.
2363 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2364 Patch provided by Eugene Bujak.
2366 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2367 accept(). Patch provided by Maxim Dounin.
2369 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2370 Patch provided by Phil Pennock.
2372 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2374 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2375 Patch provided by Brad "anomie" Jorsch.
2377 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2378 Patch provided by Dean Brooks.
2380 TK/05 Add native DKIM support (does not depend on external libraries).
2382 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2383 Patch provided by Graeme Fowler.
2385 NM/16 Bugzilla 851: Documentation example syntax fix.
2387 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2389 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2392 NM/19 Bugzilla 745: TLS version reporting.
2393 Patch provided by Phil Pennock.
2395 NM/20 Bugzilla 167: bool: condition support.
2396 Patch provided by Phil Pennock.
2398 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2399 clients. Patch provided by Phil Pennock.
2401 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2402 Patch provided by Brad "anomie" Jorsch.
2404 NM/23 Bugzilla 687: Fix misparses in eximstats.
2405 Patch provided by Heiko Schlittermann.
2407 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2408 Patch provided by Heiko Schlittermann.
2410 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2411 plus update to original patch.
2413 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2415 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2416 Patch provided by David Brownlee.
2418 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2420 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2422 NM/30 Bugzilla 888: TLS documentation bugfixes.
2424 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2426 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2427 Unlike the original bugzilla I have changed all shell scripts in src tree.
2429 NM/33 Bugzilla 898: Transport filter timeout fix.
2430 Patch by Todd Rinaldo.
2432 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2433 Patch by Serge Demonchaux.
2435 NM/35 Bugzilla 39: Base64 decode bug fixes.
2436 Patch by Jakob Hirsch.
2438 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2440 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2442 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2444 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2450 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2451 ALT-N's libdkim that I have put here:
2452 http://duncanthrax.net/exim-experimental/
2454 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2455 were called 'true' and it seems that C99 defines that as a reserved
2456 keyword to be used with 'bool' variable types. That means you could
2457 not include C99-style headers which use bools without triggering
2458 build errors in sieve.c.
2460 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2461 as mailq or other aliases. Changed the --help handling significantly
2462 to do whats expected. exim_usage() emits usage/help information.
2464 SC/01 Added the -bylocaldomain option to eximstats.
2466 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2468 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2470 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2476 PH/01 Another patch from the Sieve maintainer.
2478 PH/02 When an IPv6 address is converted to a string for single-key lookup
2479 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2480 dots are used instead of colons so that keys in lsearch files need not
2481 contain colons. This was done some time before quoting was made available
2482 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2483 (notated using the quote facility) so as to distinguish them from IPv4
2484 keys. This meant that lookups for IP addresses in host lists did not work
2485 for iplsearch lookups.
2487 This has been fixed by arranging for IPv6 addresses to be expressed with
2488 colons if the lookup type is iplsearch. This is not incompatible, because
2489 previously such lookups could never work.
2491 The situation is now rather anomalous, since one *can* have colons in
2492 ordinary lsearch keys. However, making the change in all cases is
2493 incompatible and would probably break a number of configurations.
2495 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2498 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2499 conversion specification without a maximum field width, thereby enabling
2500 a rogue spamd server to cause a buffer overflow. While nobody in their
2501 right mind would setup Exim to query an untrusted spamd server, an
2502 attacker that gains access to a server running spamd could potentially
2503 exploit this vulnerability to run arbitrary code as the Exim user.
2505 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2506 $primary_hostname instead of what libspf2 thinks the hosts name is.
2508 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2509 a directory entry by the name of the lookup key. Previously, if a
2510 symlink pointed to a non-existing file or a file in a directory that
2511 Exim lacked permissions to read, a lookup for a key matching that
2512 symlink would fail. Now it is enough that a matching directory entry
2513 exists, symlink or not. (Bugzilla 503.)
2515 PH/03 The body_linecount and body_zerocount variables are now exported in the
2518 PH/04 Added the $dnslist_matched variable.
2520 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2521 This means they are set thereafter only if the connection becomes
2524 PH/06 Added the client_condition to authenticators so that some can be skipped
2525 by clients under certain conditions.
2527 PH/07 The error message for a badly-placed control=no_multiline_responses left
2528 "_responses" off the end of the name.
2530 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2532 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2533 (without spaces) instead of just copying the configuration text.
2535 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2537 PH/11 Added $max_received_linelength.
2539 PH/12 Added +ignore_defer and +include_defer to host lists.
2541 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2542 way in which PCRE > 7.0 is built.
2544 PH/14 Implemented queue_only_load_latch.
2546 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2547 MAIL command. The effect was to mangle the value on 64-bit systems.
2549 PH/16 Another patch from the Sieve maintainer.
2551 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2553 PH/18 If a system quota error occurred while trying to create the file for
2554 a maildir delivery, the message "Mailbox is full" was not appended to the
2555 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2556 only to a quota excession during the actual writing of the file.
2558 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2559 characters?) which causes problems in log lines. The DN values are now
2560 passed through string_printing() before being added to log lines.
2562 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2563 and InterBase are left for another time.)
2565 PH/21 Added message_body_newlines option.
2567 PH/22 Guard against possible overflow in moan_check_errorcopy().
2569 PH/23 POSIX allows open() to be a macro; guard against that.
2571 PH/24 If the recipient of an error message contained an @ in the local part
2572 (suitably quoted, of course), incorrect values were put in $domain and
2573 $local_part during the evaluation of errors_copy.
2579 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2580 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2583 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2584 issue a MAIL command.
2586 PH/02 In an ACL statement such as
2588 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2590 if a client was not listed at all, or was listed with a value other than
2591 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2592 the condition was not true (as it should be), so access was not denied.
2593 The bug was that the ! inversion was incorrectly passed on to the second
2594 item. This has been fixed.
2596 PH/03 Added additional dnslists conditions == and =& which are different from
2597 = and & when the dns lookup returns more than one IP address.
2599 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2600 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2602 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2603 FSYNC, which compiles an option called disable_fsync that allows for
2604 bypassing fsync(). The documentation is heavily laced with warnings.
2606 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2608 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2609 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2610 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2611 including adding "make clean"; (3) Added -fPIC when compiling the test
2612 dynamically loaded module, to get rid of a warning.
2614 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2615 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2616 = 0s. The bug is otherwise harmless.
2618 PH/07 There was a bug in the dovecot authenticator such that the value of
2619 $auth1 could be overwritten, and so not correctly preserved, after a
2620 successful authentication. This usually meant that the value preserved by
2621 the server_setid option was incorrect.
2623 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2625 PH/09 Installed PCRE release 7.0.
2627 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2628 run for batched SMTP input. It is now run at the start of every message
2629 in the batch. While fixing this I discovered that the process information
2630 (output by running exiwhat) was not always getting set for -bs and -bS
2631 input. This is fixed, and it now also says "batched" for BSMTP.
2633 PH/11 Added control=no_pipelining.
2635 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2636 patch, slightly modified), and move the expansion of helo_data till after
2637 the connection is made in the smtp transport (so it can use these
2640 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2642 PH/14 Added log_selector = +pid.
2644 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2646 PH/16 Add ${if forany and ${if forall.
2648 PH/17 Added dsn_from option to vary the From: line in DSNs.
2650 PH/18 Flush SMTP output before performing a callout, unless control =
2651 no_callout_flush is set.
2653 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2654 was true (the default) a successful delivery failed to delete the retry
2655 item, thus causing premature timeout of the address. The bug is now
2658 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2660 PH/21 Long custom messages for fakedefer and fakereject are now split up
2661 into multiline responses in the same way that messages for "deny" and
2662 other ACL rejections are.
2664 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2665 with slight modification.
2667 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2668 draft, changing the syntax and factoring some duplicate code".
2670 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2671 for deliveries of the second and subsequent messages over the same SMTP
2674 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2675 ${reduce, with only minor "tidies".
2677 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2679 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2680 expansion side effects.
2682 PH/27 When a message times out after an over-quota error from an Exim-imposed
2683 quota, the bounce message says "mailbox is full". This message was not
2684 being given when it was a system quota that was exceeded. It now should
2687 MH/03 Made $recipients available in local_scan(). local_scan() already has
2688 better access to the recipient list through recipients_list[], but
2689 $recipients can be useful in postmaster-provided expansion strings.
2691 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2692 in the case of a MAIL command with additional options following the
2693 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2694 were accidentally chopped off.
2696 PH/29 SMTP synchronization checks are implemented when a command is read -
2697 there is a check that no more input is waiting when there shouldn't be
2698 any. However, for some commands, a delay in an ACL can mean that it is
2699 some time before the response is written. In this time, more input might
2700 arrive, invalidly. So now there are extra checks after an ACL has run for
2701 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2702 pipelining has not been advertised.
2704 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2706 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2707 returned in the SMTP response when smtp_return_error_details was false.
2708 This has been fixed.
2710 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2711 instead of the C I/O that was originally supplied, because problems were
2712 reported on Solaris.
2714 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2715 Exim which did not show up earlier: it was assuming that a call to
2716 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2717 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2718 was a macro that became an assignment, so it seemed to work. This has
2719 changed to a proper function call with a void return, hence the compile
2720 error. Exim's code has been fixed.
2722 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2725 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2727 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2729 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2730 because it turns out to be too big or there is a timeout) there is no
2731 "Completed" line in the log. When some messages of this type were
2732 selected by exigrep, they were listed as "not completed". Others were
2733 picked up by some special patterns. I have improved the selection
2734 criteria to be more general.
2736 PH/38 The host_find_failed option in the manualroute router can now be set
2737 to "ignore", to completely ignore a host whose IP address cannot be
2738 found. If all hosts are ignored, the behaviour is controlled by the new
2739 host_all_ignored option.
2741 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2742 homing or because of multiple MX records with /mx) generated more than
2743 one IP address, and the following item turned out to be the local host,
2744 all the secondary addresses of the first item were incorrectly removed
2745 from the list, along with the local host and any following hosts (which
2746 is what is supposed to happen).
2748 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2749 whoever called Exim into the -H file. In the case of the daemon it was
2750 behaving confusingly. When first started, it used values for whoever
2751 started the daemon, but after a SIGHUP it used the Exim user (because it
2752 calls itself on a restart). I have changed the code so that it now always
2755 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2756 message are rejected with the same error (e.g. no authentication or bad
2757 sender address), and a DATA command is nevertheless sent (as can happen
2758 with PIPELINING or a stupid MUA), the error message that was given to the
2759 RCPT commands is included in the rejection of the DATA command. This is
2760 intended to be helpful for MUAs that show only the final error to their
2763 PH/42 Another patch from the Sieve maintainer.
2765 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2766 Eximstats - Fixed some broken HTML links and added missing column headers
2768 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2769 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2771 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2773 PH/43 Yet another patch from the Sieve maintainer.
2775 PH/44 I found a way to check for a TCP/IP connection going away before sending
2776 the response to the final '.' that terminates a message, but only in the
2777 case where the client has not sent further data following the '.'
2778 (unfortunately, this is allowed). However, in many cases there won't be
2779 any further data because there won't be any more messages to send. A call
2780 to select() can be used: if it shows that the input is "ready", there is
2781 either input waiting, or the socket has been closed. An attempt to read
2782 the next input character can distinguish the two cases. Previously, Exim
2783 would have sent an OK response which the client would never have see.
2784 This could lead to message repetition. This fix should cure that, at
2785 least in a lot of common cases.
2787 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2788 advertised in response to EHLO.
2794 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2795 fixed by 4.65/MH/01 (is this a record?) are fixed:
2797 (i) An empty string was always treated as zero by the numeric comparison
2798 operators. This behaviour has been restored.
2800 (ii) It is documented that the numeric comparison operators always treat
2801 their arguments as decimal numbers. This was broken in that numbers
2802 starting with 0 were being interpreted as octal.
2804 While fixing these problems I realized that there was another issue that
2805 hadn't been noticed. Values of message_size_limit (both the global option
2806 and the transport option) were treated as octal if they started with 0.
2807 The documentation was vague. These values are now always treated as
2808 decimal, and I will make that clear in the documentation.
2814 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2815 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2818 MH/01 Don't check that the operands of numeric comparison operators are
2819 integers when their expansion is in "skipping" mode (fixes bug
2820 introduced by 4.64-PH/07).
2822 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2823 child addresses, Exim now panics and dies. Previously, because the count
2824 is held in a short int, deliveries were likely to be lost. As such a
2825 large number of recipients for a single message is ridiculous
2826 (performance will be very, very poor), I have chosen to impose a limit
2827 rather than extend the field.
2833 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2834 leftover -K file (the existence of which was triggered by #402).
2835 While we were at it, introduced process PID as part of the -K
2836 filename. This should rule out race conditions when creating
2839 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2840 processing considerably. Previous code took too long for large mails,
2841 triggering a timeout which in turn triggers #401.
2843 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2844 in the DK code in transports.c. sendfile() is not really portable,
2845 hence the _LINUX specificness.
2847 TF/01 In the add_headers option to the mail command in an Exim filter,
2848 there was a bug that Exim would claim a syntax error in any
2849 header after the first one which had an odd number of characters
2852 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2853 callout verification, Exim cached a "reject" for the entire domain. This
2854 is correct for most verifications, but it is not correct for a recipient
2855 verification with use_sender or use_postmaster set, because in that case
2856 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2857 case of MAIL FROM:<> rejection from other early rejections (e.g.
2858 rejection of HELO). When verifying a recipient using a non-null MAIL
2859 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2860 Whatever the result of the callout, the value of the domain cache is
2861 left unchanged (for any other kind of callout, getting as far as trying
2862 RCPT means that the domain itself is ok).
2864 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2867 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2868 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2869 session, the daemon ignores ECONNECT errors and logs others; it now
2870 ignores EPIPE as well.
2872 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2873 (quoted-printable decoding).
2875 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2876 later the small subsequent patch to fix an introduced bug.
2878 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2880 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2882 PH/08 An error is now given if message_size_limit is specified negative.
2884 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2885 to be given (somewhat) arbitrary names.
2887 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2890 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2891 miscellaneous code fixes
2893 PH/10 Added the log_reject_target ACL modifier to specify where to log
2896 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2897 hostname. This is wrong, because it relates to the incoming message (and
2898 probably the interface on which it is arriving) and not to the outgoing
2899 callout (which could be using a different interface). This has been
2900 changed to use the value of the helo_data option from the smtp transport
2901 instead - this is what is used when a message is actually being sent. If
2902 there is no remote transport (possible with a router that sets up host
2903 addresses), $smtp_active_hostname is used.
2905 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2906 tweaks were necessary in order to get it to work (see also 21 below):
2907 (a) The code assumed that strncpy() returns a negative number on buffer
2908 overflow, which isn't the case. Replaced with Exim's string_format()
2910 (b) There were several signed/unsigned issues. I just did the minimum
2911 hacking in of casts. There is scope for a larger refactoring.
2912 (c) The code used strcasecmp() which is not a standard C function.
2913 Replaced with Exim's strcmpic() function.
2914 (d) The code set only $1; it now sets $auth1 as well.
2915 (e) A simple test gave the error "authentication client didn't specify
2916 service in request". It would seem that Dovecot has changed its
2917 interface. Fortunately there's a specification; I followed it and
2918 changed what the client sends and it appears to be working now.
2920 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2923 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2924 address A is aliased to B and C, where B exists and C does not. Without
2925 -v the output is "A verified" because verification stops after a
2926 successful redirection if more than one address is generated. However,
2927 with -v the child addresses are also verified. Exim was outputting "A
2928 failed to verify" and then showing the successful verification for C,
2929 with its parentage. It now outputs "B failed to verify", showing B's
2930 parentage before showing the successful verification of C.
2932 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2933 look up a TXT record in a specific list after matching in a combined
2936 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2937 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2938 they consult the DNS. I had assumed they would set it the way they
2939 wanted; and indeed my experiments on Linux seem to show that in some
2940 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2941 To be on the safe side, however, I have now made the interface to
2942 host_find_byname() similar to host_find_bydns(), with an argument
2943 containing the DNS resolver options. The host_find_byname() function now
2944 sets these options at its start, just as host_find_bydns() does. The smtp
2945 transport options dns_qualify_single and dns_search_parents are passed to
2946 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2947 of host_find_byname() use the default settings of RES_DEFNAMES
2948 (qualify_single) but not RES_DNSRCH (search_parents).
2950 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2951 spool_read_header() do less string testing, by means of a preliminary
2952 switch on the second character of optional "-foo" lines. (This is
2953 overdue, caused by the large number of possibilities that now exist.
2954 Originally there were few.) While I was there, I also converted the
2955 str(n)cmp tests so they don't re-test the leading "-" and the first
2956 character, in the hope this might squeeze out yet more improvement.
2958 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2959 flag allowing group syntax was set by the header_syntax check but not
2960 turned off, possible causing trouble later; (2) The flag was not being
2961 set at all for the header_verify test, causing "group"-style headers to
2962 be rejected. I have now set it in this case, and also caused header_
2963 verify to ignore an empty address taken from a group. While doing this, I
2964 came across some other cases where the code for allowing group syntax
2965 while scanning a header line wasn't quite right (mostly, not resetting
2966 the flag correctly in the right place). These bugs could have caused
2967 trouble for malformed header lines. I hope it is now all correct.
2969 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2970 with the "reply" argument non-NULL. The code, however (which originally
2971 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2972 but it didn't always do it. This confused somebody who was copying the
2973 code for some other use. I have removed all the tests.
2975 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2976 feature that was used to support insecure browsers during the U.S. crypto
2977 embargo. It requires special client support, and Exim is probably the
2978 only MTA that supported it -- and would never use it because real RSA is
2979 always available. This code has been removed, because it had the bad
2980 effect of slowing Exim down by computing (never used) parameters for the
2981 RSA_EXPORT functionality.
2983 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2984 authenticator to fail if there's a tab character in the incoming data
2985 (there should never be unless someone is messing about, as it's supposed
2986 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2987 if the connection is using TLS or if the remote IP is the same as the
2988 local IP, and the "valid-client-cert option" if a client certificate has
2991 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2992 authenticators. This can be used for authorization after authentication
2993 succeeds. (In the case of plaintext, it servers for both authentication
2996 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2997 if any retry times were supplied.
2999 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3000 connection, where there is no client IP address to check. In this
3001 situation, the verify now always succeeds.
3003 PH/25 Applied John Jetmore's -Mset patch.
3005 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3007 PH/27 In a string expansion for a processed (not raw) header when multiple
3008 headers of the same name were present, leading whitespace was being
3009 removed from all of them, but trailing whitespace was being removed only
3010 from the last one. Now trailing whitespace is removed from each header
3011 before concatenation. Completely empty headers in a concatenation (as
3012 before) are ignored.
3014 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3015 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3017 PH/29 [Removed. This was a change that I later backed out, and forgot to
3018 correct the ChangeLog entry (that I had efficiently created) before
3019 committing the later change.]
3021 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3022 address errors (4xx response to RCPT) over the same connection as other
3023 messages routed to the same hosts. Such deliveries are always "forced",
3024 so retry times are not inspected. This resulted in far too many retries
3025 for the affected addresses. The effect occurred only when there were more
3026 hosts than the hosts_max_try setting in the smtp transport when it had
3027 the 4xx errors. Those hosts that it had tried were not added to the list
3028 of hosts for which the message was waiting, so if all were tried, there
3029 was no problem. Two fixes have been applied:
3031 (i) If there are any address or message errors in an SMTP delivery, none
3032 of the hosts (tried or untried) are now added to the list of hosts
3033 for which the message is waiting, so the message should not be a
3034 candidate for sending over the same connection that was used for a
3035 successful delivery of some other message. This seems entirely
3036 reasonable: after all the message is NOT "waiting for some host".
3037 This is so "obvious" that I'm not sure why it wasn't done
3038 previously. Hope I haven't missed anything, but it can't do any
3039 harm, as the worst effect is to miss an optimization.
3041 (ii) If, despite (i), such a delivery is accidentally attempted, the
3042 routing retry time is respected, so at least it doesn't keep
3043 hammering the server.
3045 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3046 in ${readsocket because some servers need this prod.
3048 PH/32 Added some extra debug output when updating a wait-xxx database.
3050 PH/33 The hint "could be header name not terminated by colon", which has been
3051 given for certain expansion errors for a long time, was not being given
3052 for the ${if def:h_colon_omitted{... case.
3054 PH/34 The spec says: "With one important exception, whenever a domain list is
3055 being scanned, $domain contains the subject domain." There was at least
3056 one case where this was not true.
3058 PH/35 The error "getsockname() failed: connection reset by peer" was being
3059 written to the panic log as well as the main log, but it isn't really
3060 panic-worthy as it just means the connection died rather early on. I have
3061 removed the panic log writing for the ECONNRESET error when getsockname()
3064 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3065 runs only) independently of the message's sender address. This meant
3066 that, if the 4xx error was in fact related to the sender, a different
3067 message to the same recipient with a different sender could confuse
3068 things. In particular, this can happen when sending to a greylisting
3069 server, but other circumstances could also provoke similar problems.
3070 I have changed the default so that the retry time for these errors is now
3071 based a combination of the sender and recipient addresses. This change
3072 can be overridden by setting address_retry_include_sender=false in the
3075 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3076 remote server are returned as part of bounce messages. This was not
3077 happening for LMTP over a pipe (the lmtp transport), but now it is the
3078 same for both kinds of LMTP.
3080 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3081 in header lines that were in fact CNAMEs. This is no longer the case.
3083 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3084 and queue runs started by the daemon processed all messages. This has
3085 been fixed so that -R and -S can now usefully be given with -q<time>.
3087 PH/40 Import PCRE release 6.7 (fixes some bugs).
3089 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3091 PH/42 Give an error if -q is specified more than once.
3093 PH/43 Renamed the variables $interface_address and $interface_port as
3094 $received_ip_address and $received_port, to make it clear that these
3095 values apply to message reception, and not to the outgoing interface when
3096 a message is delivered. (The old names remain recognized, of course.)
3098 PH/44 There was no timeout on the connect() call when using a Unix domain
3099 socket in the ${readsocket expansion. There now is.
3101 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3102 be meaningful with "accept".
3104 SC/01 Eximstats V1.43
3105 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3107 SC/02 Eximstats V1.44
3108 Use a glob alias rather than an array ref in the generated
3109 parser. This improves both readability and performance.
3111 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3112 Collect SpamAssassin and rejection statistics.
3113 Don't display local sender or destination tables unless
3114 there is data to show.
3115 Added average volumes into the top table text output.
3117 SC/04 Eximstats V1.46
3118 Collect data on the number of addresses (recipients)
3119 as well as the number of messages.
3121 SC/05 Eximstats V1.47
3122 Added 'Message too big' to the list of mail rejection
3123 reasons (thanks to Marco Gaiarin).
3125 SC/06 Eximstats V1.48
3126 Mainlog lines which have GMT offsets and are too short to
3127 have a flag are now skipped.
3129 SC/07 Eximstats V1.49 (Alain Williams)
3130 Added the -emptyok flag.
3132 SC/08 Eximstats V1.50
3133 Fixes for obtaining the IP address from reject messages.
3135 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3136 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3137 whitespace changes from 4.64-PH/27
3139 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3142 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3143 are found, allow negative numbers in numeric criteria)
3145 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3147 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3148 to match changes made in 4.64-PH/43
3150 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3152 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3153 "domain" is actually an IP address, and force a failure. This locks out
3154 those revolvers/nameservers that support "A-for-A" lookups, in
3155 contravention of the specifications.
3157 PH/48 When a host name was looked up from an IP address, and the subsequent
3158 forward lookup of the name timed out, the host name was left in
3159 $sender_host_name, contrary to the specification.
3161 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3162 restricted to single-key lookups, Exim was not diagnosing an error if
3163 * or *@ was used with a query-style lookup.
3165 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3167 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3168 long ago, but noone interested enough thought of it. Let's just say that
3169 the "1.1" means that there are some new functions that weren't there at
3170 some point in the past.
3172 PH/51 Error processing for expansion failure of helo_data from an smtp
3173 transport during callout processing was broken.
3175 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3176 tested/used via the -bh/-bhc/-bs options.
3178 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3179 bug, fixed in subsequent PCRE releases).
3181 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3182 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3184 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3190 SC/01 Use a glob alias rather than an array ref in eximstats generated
3191 parser. This improves both readability and performance.
3193 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3194 Don't display local sender or destination tables in eximstats unless
3195 there is data to show.
3196 Added average volumes into the eximstats top table text output.
3198 SC/03 Collect data on the number of addresses (recipients) as well
3199 as the number of messages in eximstats.
3201 TF/01 Correct an error in the documentation for the redirect router. Exim
3202 does (usually) call initgroups() when daemonizing.
3204 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3205 with consistent privilege compared to when running as a daemon.
3207 TF/03 Note in the spec that $authenticated_id is not set for local
3208 submissions from trusted users.
3210 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3211 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3213 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3214 by adding some example configuration directives to the default
3215 configuration file. A little bit of work is required to uncomment the
3216 directives and define how usernames and passwords are checked, but
3217 there is now a framework to start from.
3219 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3220 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3221 without this. I don't know how relevant this is to other LDAP libraries.
3223 PH/02 Add the verb name to the "unknown ACL verb" error.
3225 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3227 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3229 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3230 directory not expanded when it should be if an expanded home directory
3231 was set for the address (which is overridden by the transport).
3233 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3236 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3237 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3238 because it is too late at that time, and has no effect.
3240 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3241 security issue with \' (bugzilla #107). I could not use the
3242 PQescapeStringConn() function, because it needs a PGconn value as one of
3245 PH/08 When testing addresses using -bt, indicate those final addresses that
3246 are duplicates that would not cause an additional delivery. At least one
3247 person was confused, thinking that -bt output corresponded to deliveries.
3248 (Suppressing duplicates isn't a good idea as you lose the information
3249 about possibly different redirections that led to the duplicates.)
3251 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3252 systems where poll() doesn't work, in particular OS X.
3254 PH/10 Added more information to debugging output for retry time not reached.
3256 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3257 operations in malware.c.
3259 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3262 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3263 syslog_duplication set false, log lines that would normally be written
3264 both the the main log and to the reject log were not written to syslog at
3267 PH/14 In the default configuration, change the use of "message" in ACL warn
3268 statements to "add_header".
3270 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3271 not followed by a command (e.g. "seen endif").
3273 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3274 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3277 PH/17 Added extra conditions to the default value of delay_warning_condition
3281 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3282 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3283 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3286 The Auto-Submitted: and various List- headers are standardised, whereas I
3287 don't think Precedence: ever was.
3289 PH/18 Refactored debugging code in route_finduser() to show more information,
3290 in particular, the error code if getpwnam() issues one.
3292 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3293 This is apparently needed in addition to the PH/07 change above to avoid
3294 any possible encoding problems.
3296 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3297 but not after initializing Perl.
3299 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3300 output them only if debugging. By default they are written stderr,
3301 apparently, which is not desirable.
3303 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3306 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3309 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3311 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3312 authenticated or an ident call has been made. Suppress the default
3313 values for $authenticated_id and $authenticated_sender (but permit -oMai
3314 and -oMas) when testing with -bh.
3316 PH/24 Re-jigged the order of the tests in the default configuration so that the
3317 tests for valid domains and recipients precede the DNS black list and CSA
3318 tests, on the grounds that those ones are more expensive.
3320 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3321 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3322 This bug exists in every version of Exim that I still have, right back to
3325 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3326 However, an attempt to turn on case-sensitivity in a regex key by
3327 including (?-i) didn't work because the subject string was already
3328 lowercased, and the effects were non-intuitive. It turns out that a
3329 one-line patch can be used to allow (?-i) to work as expected.
3335 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3336 other effects) broke the use of negated acl sub-conditions.
3338 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3341 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3342 "Deny" causes Exim to reject the incoming connection with a 554 error.
3343 Unfortunately, if there is a major crisis, such as a disk failure,
3344 tcp-wrappers gives "deny", whereas what one would like would be some
3345 kind of temporary error. A kludge has been added to help with this.
3346 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3347 554 error is used if errno is still zero or contains ENOENT (which occurs
3348 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3351 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3353 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3354 errors. Otherwise a message that provokes a temporary error (when other
3355 messages do not) can cause a whole host to time out.
3357 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3358 addresses were routed directly to files or pipes from a redirect router.
3359 File deliveries just didn't batch; pipe deliveries might have suffered
3362 PH/06 A failure to get a lock for a hints database would erroneously always say
3363 "Failed to get write lock", even when it was really a read lock.
3365 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3366 of 0600. This has been changed to use the value of the lockfile_mode
3367 option (which defaults to 0600).
3369 PH/08 Applied small patch from the Sieve maintainer.
3371 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3372 folder from quota calculations, a direct delivery into this folder messed
3373 up the contents of the maildirsize file. This was because the regex was
3374 used only to exclude .Trash (or whatever) when the size of the mailbox
3375 was calculated. There was no check that a delivery was happening into an
3376 excluded directory. This bug has been fixed by ignoring all quota
3377 processing for deliveries into excluded directories.
3379 PH/10 Added the maildirfolder_create_regex option to appendfile.
3385 PH/01 The code for finding all the local interface addresses on a FreeBSD
3386 system running IPv6 was broken. This may well have applied to all BSD
3387 systems, as well as to others that have similar system calls. The broken
3388 code found IPv4 interfaces correctly, but gave incorrect values for the
3389 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3390 that it would not match correctly against @[] and not recognize the IPv6
3393 PH/02 The ipliteral router was not recognizing addresses of the form user@
3394 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3396 PH/03 Added disable_ipv6.
3398 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3399 decoded form, because it is most often used to construct To: headers
3400 lines in autoreplies, and the decoded form may well be syntactically
3401 invalid. However, $reply_address has leading white space removed, and all
3402 newlines turned into spaces so that the autoreply transport does not
3405 PH/05 If group was specified without a user on a router, and no group or user
3406 was specified on a transport, the group from the router was ignored.
3408 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3409 for visible compile-time settings that can be used to change these
3410 numbers, for those that want even more. Backwards compatibility with old
3411 spool files has been maintained. However, going back to a previous Exim
3412 release will lost any variables that are in spool files.
3414 PH/07 Two small changes when running in the test harness: increase delay when
3415 passing a TCP/IP connection to a new process, in case the original
3416 process has to generate a bounce, and remove special handling of
3417 127.0.0.2 (sic), which is no longer necessary.
3419 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3420 be the same on different OS.
3422 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3425 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3426 whether --show-vars was specified or not
3428 JJ/02 exipick: Added support for new ACL variable spool format introduced
3431 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3432 syntactically invalid From: or Reply-to: line, and a filter used this to
3433 generate an autoreply, and therefore failed to obtain an address for the
3434 autoreply, Exim could try to deliver to a non-existent relative file
3435 name, causing unrelated and misleading errors. What now happens is that
3436 it logs this as a hard delivery error, but does not attempt to create a
3439 PH/11 The exinext utility has a -C option for testing purposes, but although
3440 the given file was scanned by exinext itself; it wasn't being passed on
3441 when Exim was called.
3443 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3444 an end-of-file indication when reading a command response.
3446 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3447 compiled. In many other places in Exim, IPv6 addresses are always
3448 recognized, so I have changed this. It also means that IPv4 domain
3449 literals of the form [IPV4:n.n.n.n] are now always recognized.
3451 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3452 used if the router is not running as root, for example, when verifying at
3453 ACL time, or when using -bh. The debugging output from this situation was
3454 non-existent - all you got was a failure to exec. I have made two
3457 (a) Failures to set uid/gid, the current directory, or a process leader
3458 in a subprocess such as that created by queryprogram now generate
3459 suitable debugging output when -d is set.
3461 (b) The queryprogram router detects when it is not running as root,
3462 outputs suitable debugging information if -d is set, and then runs
3463 the subprocess without attempting to change uid/gid.
3465 PH/15 Minor change to Makefile for building test_host (undocumented testing
3468 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3469 additional section of a DNS packet that returns MX or SRV records.
3470 Instead, it always explicitly searches for A/AAAA records. This avoids
3471 major problems that occur when a DNS server includes only records of one
3472 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3473 fixed another bug: if SRV records were looked up and the corresponding
3474 address records were *not* found in the additional section, the port
3475 values from the SRV records were lost.
3477 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3478 using the correct key (the original address) when searching the retry
3479 rules in order to find which one to use for generating the retry hint.
3481 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3482 adding the default one. Similarly, if it contains a Reply-To: header, the
3483 errors_reply_to option, if set, is not used.
3485 PH/19 When calculating a retry time, Exim used to measure the "time since
3486 failure" by looking at the "first failed" field in the retry record. Now
3487 it does not use this if it is later than than the arrival time of the
3488 message. Instead it uses the arrival time. This makes for better
3489 behaviour in cases where some deliveries succeed, thus re-setting the
3490 "first failed" field. An example is a quota failure for a huge message
3491 when small messages continue to be delivered. Without this change, the
3492 "time since failure" will always be short, possible causing more frequent
3493 delivery attempts for the huge message than are intended.
3494 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3496 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3497 $1, $2, $3) because the numerical variables can be reset during some
3498 expansion items (e.g. "match"), thereby losing the authentication data.
3500 PH/21 Make -bV show the size of off_t variables so that the test suite can
3501 decide whether to run tests for quotas > 2G.
3503 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3504 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3505 filecount value is greater than 2G or if a quota value is greater than 2G
3506 on a system where the size of off_t is not greater than 4, a panic error
3509 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3510 never match. The debug and -bh output now contains an explicit error
3511 message indicating a malformed IPv4 address or mask.
3513 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3514 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3515 PH/23 above applies.
3517 PH/25 Do not write to syslog when running in the test harness. The only
3518 occasion when this arises is a failure to open the main or panic logs
3519 (for which there is an explicit test).
3521 PH/26 Added the /no_tell option to "control=freeze".
3523 PH/27 If a host name lookup failed very early in a connection, for example, if
3524 the IP address matched host_lookup and the reverse lookup yielded a name
3525 that did not have a forward lookup, an error message of the form "no IP
3526 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3527 could be logged. Now it outputs the IP address instead of "NULL".
3529 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3530 allows the sender and the authenticated sender to be set when
3531 submitting a message from within Exim. Since child_open_exim() is
3532 documented for local_scan(), the new function should be too.
3534 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3535 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3536 results in an empty string is now treated as unset.
3538 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3540 PH/31 Added sender_verify_fail logging option.
3542 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3543 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3544 tidied the source and removed it altogether.
3546 PH/33 When a queue run was abandoned because the load average was too high, a
3547 log line was always written; now it is written only if the queue_run log
3548 selector is set. In addition, the log line for abandonment now contains
3549 information about the queue run such as the pid. This is always present
3550 in "start" and "stop" lines but was omitted from the "abandon" line.
3552 PH/34 Omit spaces between a header name and the colon in the error message that
3553 is given when verify = headers_syntax fails (if there are lots of them,
3554 the message gets confusing).
3556 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3557 names, as there are now some PTR records that contain slashes. This check
3558 is only to protect against broken name servers that fall over on strange
3559 characters, so the fact that it applies to all lookups doesn't matter.
3561 PH/36 Now that the new test suite is complete, we can remove some of the
3562 special code in Exim that was needed for the old test suite. For example,
3563 sorting DNS records because real resolvers return them in an arbitrary
3564 order. The new test suite's fake resolver always returns records in the
3567 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3568 bounces) except when queue_only is set, to avoid logging races between
3569 the different processes.
3571 PH/38 Panic-die if .include specifies a non-absolute path.
3573 PH/39 A tweak to the "H" retry rule from its user.
3575 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3576 a label. They prevented compilation on older perls.
3578 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3579 a warning to be raised on newish perls.
3581 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3582 on queue. Changes to match documented behaviour of showing count of
3583 messages matching specified criteria.
3585 PH/40 Changed the default ident timeout from 30s to 5s.
3587 PH/41 Added support for the use of login_cap features, on those BSD systems
3588 that have them, for controlling the resources used by pipe deliveries.
3590 PH/42 The content-scanning code uses fopen() to create files in which to put
3591 message data. Previously it was not paying any attention to the mode of
3592 the files. Exim runs with umask(0) because the rest of the code creates
3593 files with open(), and sets the required mode explicitly. Thus, these
3594 files were ending up world-writeable. This was not a big issue, because,
3595 being within the spool directory, they were not world-accessible. I have
3596 created a function called modefopen, which takes an additional mode
3597 argument. It sets umask(777), creates the file, chmods it to the required
3598 mode, then resets the umask. All the relevant calls to fopen() in the
3599 content scanning code have been changed to use this function.
3601 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3602 to 24 hours. This avoids potential overflow problems when processing G
3603 and H retry rules. I suspect nobody ever tinkers with this value.
3605 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3607 PH/45 When the plaintext authenticator is running as a client, the server's
3608 challenges are checked to ensure they are valid base64 strings. By
3609 default, the authentication attempt is cancelled if an invalid string is
3610 received. Setting client_ignore_invalid_base64 true ignores these errors.
3611 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3612 they are received. Thus, the responses can be made to depend on the
3613 challenges. If an invalid string is ignored, an empty string is placed in
3616 PH/46 Messages that are created by the autoreply transport now contains a
3617 References: header, in accordance with RFCs 2822 and 3834.
3619 PH/47 Added authenticated_sender_force to the smtp transport.
3621 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3623 PH/49 Installed latest patch from the Sieve maintainer.
3625 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3626 was also set, the appendfile transport was unnecessarily scanning a
3627 directory of message files (e.g. for maildir delivery) to find the count
3628 of files (along with the size), even though it did not need this
3629 information. It now does the scan only if it needs to find either the
3630 size of the count of files.
3632 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3634 PH/52 Two bugs concerned with error handling when the smtp transport is
3637 (i) Exim was not creating retry information for temporary errors given
3638 for individual recipients after the DATA command when the smtp transport
3639 was used in LMTP mode. This meant that they could be retried too
3640 frequently, and not timed out correctly.
3642 (ii) Exim was setting the flag that allows error details to be returned
3643 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3644 recipients that were returned after the DATA command.
3646 PH/53 This is related to PH/52, but is more general: for any failing address,
3647 when detailed error information was permitted to be returned to the
3648 sender, but the error was temporary, then after the final timeout, only
3649 "retry timeout exceeded" was returned. Now it returns the full error as
3650 well as "retry timeout exceeded".
3652 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3653 do this, and (what is worse) MTAs that accept it.
3655 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3656 will now be deprecated.
3658 PH/56 New os.c-cygwin from the Cygwin maintainer.
3660 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3661 formats (previously only available in exim formats via -bpr, -bpru,
3662 and -bpra. Now also available in native and exiqgrep formats)
3664 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3665 with very large, slow to parse queues
3667 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3669 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3671 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3672 responses to authentication challenges, though it was showing the
3673 challenges; (ii) I've removed the CR characters from the debug output for
3676 PH/58 Allow for the insertion of a newline as well as a space when a string
3677 is turned into more than one encoded-word during RFC 2047 encoding. The
3678 Sieve code now uses this.
3680 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3681 data_4xx, lost_connection, tls_required.
3683 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3684 message was being sent as an SMTP response.
3686 PH/61 Add -l and -k options to exicyclog.
3688 PH/62 When verifying, if an address was redirected to one new address, so that
3689 verification continued, and the new address failed or deferred after
3690 having set something in $address_data, the value of $address_data was not
3691 passed back to the ACL. This was different to the case when no
3692 redirection occurred. The value is now passed back in both cases.
3694 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3695 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3696 use login_cap.h, so on its own it isn't the distinguishing feature. The
3697 new name refers directly to the setclassresources() function.
3699 PH/65 Added configuration files for NetBSD3.
3701 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3703 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3704 is preferred over IPv4.
3706 PH/68 The bounce_return_message and bounce_return_body options were not being
3707 honoured for bounces generated during the reception of non-SMTP messages.
3708 In particular, this applied to messages rejected by the ACL. This bug has
3709 been fixed. However, if bounce_return_message is true and bounce_return_
3710 body is false, the headers that are returned for a non-SMTP message
3711 include only those that have been read before the error was detected.
3712 (In the case of an ACL rejection, they have all been read.)
3714 PH/69 The HTML version of the specification is now built in a directory called
3715 spec_html instead of spec.html, because the latter looks like a path with
3716 a MIME-type, and this confuses some software.
3718 PH/70 Catch two compiler warnings in sieve.c.
3720 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3721 function verify_get_ident() calls ip_connect() to connect a socket, but
3722 if the "connect()" function timed out, ip_connect() used to close the
3723 socket. However, verify_get_ident() also closes the socket later, and in
3724 between Exim writes to the log, which may get opened at this point. When
3725 the socket was closed in ip_connect(), the log could get the same file
3726 descriptor number as the socket. This naturally causes chaos. The fix is
3727 not to close the socket in ip_connect(); the socket should be closed by
3728 the function that creates it. There was only one place in the code where
3729 this was missing, in the iplookup router, which I don't think anybody now
3730 uses, but I've fixed it anyway.
3732 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3733 well as to direct DNS lookups. Otherwise the handling of names in host
3734 lists is inconsistent and therefore confusing.
3740 PH/01 Two changes to the default runtime configuration:
3742 (1) Move the checks for relay_from_hosts and authenticated clients from
3743 after to before the (commented out) DNS black list checks.
3745 (2) Add control=submission to the relay_from_hosts and authenticated
3746 clients checks, on the grounds that messages accepted by these
3747 statements are most likely to be submissions.
3749 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3751 (1) Generate an error if the third argument for the ${prvs expansion is
3754 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3757 (3) Reset the variables that are obtained from the first argument of
3758 ${prvscheck and used in the second argument before leaving the code,
3759 because their memory is reclaimed, so using them afterwards may do
3762 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3763 one (it's much easier than Tom thought :-).
3765 (5) Because of (4), we can now allow for the use of $prvscheck_result
3766 inside the third argument.
3768 PH/03 For some reason, the default setting of PATH when running a command from
3769 a pipe transport was just "/usr/bin". I have changed it to
3772 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3773 anything to be listed in the output from -bV.
3775 PH/05 When a filter generated an autoreply, the entire To: header line was
3776 quoted in the delivery log line, like this:
3778 => >A.N.Other <ano@some.domain> <original@ddress> ...
3780 This has been changed so that it extracts the operative address. There
3781 may be more than one such address. If so, they are comma-separated, like
3784 => >ano@some.domain,ona@other.domain <original@ddress> ...
3786 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3787 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3788 not being looked up in the rDNS to get a host name, Exim was showing the
3789 IP address twice in Received: lines, even though the IP addresses were
3790 identical. For example:
3792 Received: from [1.2.3.4] (helo=[1.2.3.4])
3794 However, if the real host name was known, it was omitting the HELO data
3795 if it matched the actual IP address. This has been tidied up so that it
3796 doesn't show the same IP address twice.
3798 PH/07 When both +timestamp and +memory debugging was on, the value given by
3799 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3800 called by the debug printing, thereby overwriting the timestamp buffer.
3801 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3803 PH/08 When the original message was included in an autoreply transport, it
3804 always said "this is a copy of the message, including all the headers",
3805 even if body_only or headers_only was set. It now gives an appropriate
3808 PH/09 Applied a patch from the Sieve maintainer which:
3810 o fixes some comments
3811 o adds the (disabled) notify extension core
3812 o adds some debug output for the result of if/elsif tests
3813 o points to the current vacation draft in the documentation
3814 and documents the missing references header update
3818 o fixes a bug in processing the envelope test (when testing
3819 multiple envelope elements, the last element determined the
3822 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3823 Electronic Mail") by including:
3825 Auto-submitted: auto-generated
3827 in the messages that it generates (bounce messages and others, such as
3828 warnings). In the case of bounce messages for non-SMTP messages, there was
3829 also a typo: it was using "Auto_submitted" (underscore instead of
3830 hyphen). Since every message generated by Exim is necessarily in response
3831 to another message, thes have all been changed to:
3833 Auto-Submitted: auto-replied
3835 in accordance with these statements in the RFC:
3837 The auto-replied keyword:
3839 - SHOULD be used on messages sent in direct response to another
3840 message by an automatic process,
3842 - MUST NOT be used on manually-generated messages,
3844 - MAY be used on Delivery Status Notifications (DSNs) and Message
3845 Disposition Notifications (MDNs),
3847 - MUST NOT be used on messages generated by automatic or periodic
3848 processes, except for messages which are automatic responses to
3851 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3852 to the default Received: header definition.
3854 PH/12 Added log selector acl_warn_skipped (default on).
3856 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3857 variables because (a) they are in the wrong storage pool and (b) even if
3858 they were copied, it wouldn't work properly because of the caching.
3860 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3861 checking when decoding. Apparently there are clients that generate
3862 overlong encoded strings. Why am I not surprised?
3864 PH/15 If the first argument of "${if match_address" was not empty, but did not
3865 contain an "@" character, Exim crashed. Now it writes a panic log message
3866 and treats the condition as false.
3868 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3870 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3871 extension "envelope-auth". The code is finished and in agreement with
3872 other implementations, but there is no documentation so far and in fact,
3873 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3874 not changing the active code.
3876 Print executed "if" and "elsif" statements when debugging is used. This
3877 helps a great deal to understand what a filter does.
3879 Document more things not specified clearly in RFC3028. I had all this
3880 sorted out, when out of a sudden new issues came to my mind. Oops."
3882 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3885 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3886 canonical form (as documented). However, after a host name lookup from
3887 the IP address, check_host() was doing a simple string comparison with
3888 addresses acquired from the DNS when checking that the found name did
3889 have the original IP as one of its addresses. Since any found IPv6
3890 addresses are likely to be in abbreviated form, the comparison could
3891 fail. Luckily, there already exists a function for doing the comparison
3892 by converting both addresses to binary, so now that is used instead of
3893 the text comparison.
3895 PH/20 There was another similar case to PH/19, when a complete host name was
3896 given in a host list; looking up its IP address could give an abbreviated
3897 form, whereas the current host's name might or might not be abbreviated.
3898 The same fix has been applied.
3904 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3905 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3908 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3909 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3911 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3913 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3914 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3915 Also, for "Auto-Submitted", treat anything other than "no" as
3916 non-personal, in accordance with RFC 3834. (Previously it treated
3917 anything starting "auto-" as non-personal.)
3919 TF/01 The control=submission/name=... option had a problem with syntax
3920 errors if the name included a slash character. The /name= option
3921 now slurps the rest of the string, so it can include any characters
3922 but it must come last in the list of options (after /sender_retain
3925 PH/05 Some modifications to the interface to the fake nameserver for the new
3933 TK/01 Added the "success_on_redirect" address verification option. See
3934 NewStuff for rationale and an example.
3936 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3938 PH/02 Patch to exigrep to allow it to work on syslog lines.
3940 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3941 fread() to skip over the body file's header line, because in Cygwin the
3942 header line is locked and is inaccessible.
3944 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3945 co-exist for some time) to make it clear that it is the Exim ID that is
3946 referenced, not the Message-ID: header line.
3948 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3949 string_format() function, because snprintf() does not exist on all
3952 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3953 ${stat: expansion item.
3955 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3956 protocol synchronization error", to keep the pedants happy.
3958 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3959 well as for IRIX systems, when gcc is being used. See the host.c source
3962 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3964 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3967 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3968 transport and to the smtp transport in LMTP mode.
3970 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3972 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3973 run a filter in a subprocess. This could lead to confusion in subsequent
3974 lookups in the parent process. There should also be a search_tidyup() at
3975 the end of the subprocess.
3977 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3978 only if the host matched helo_try_verify_hosts, which caused the
3979 verification to occur when the EHLO/HELO command was issued. The ACL just
3980 tested the remembered result. Now, if a previous verification attempt has
3981 not happened, "verify = helo" does it there and then.
3983 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3985 TK/03 Fix log output including CR from clamd.
3987 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3988 exist provoked a memory error which could cause a segfault.
3990 PH/15 Installed PCRE 6.2
3992 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3994 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3995 of the problem. Specifically, suggested +O2 rather than +O1 for the
3998 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4000 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4001 record, the port was not correctly logged when the outgoing_port log
4002 selector was set (it logged the transort's default port).
4004 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4005 fallback_hosts, and "hosts" in the smtp transport.
4007 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4008 host errors such as "Connection refused".
4010 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4011 authentication with radiusclient 0.4.9:
4013 - Error returned from rc_read_config was caught wrongly
4014 - Username/password not passed on to radius server due to wrong length.
4016 The presumption is that some radiusclient API changes for 4.51/PH/17
4017 were not taken care of correctly. The code is still untested by me (my
4018 Linux distribution still has 0.3.2 of radiusclient), but it was
4019 contributed by a Radius user.
4021 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4022 expanding the "port" option of the smtp transport.
4024 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4025 while reading a MIME header. Thanks to Tom Hughes for a patch.
4027 PH/24 Include config.h inside local_scan.h so that configuration settings are
4030 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4031 that in an ACL for RCPT (for example), you can examine exactly what was
4034 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4035 commands, but it was not correctly comparing the address with the actual
4036 client host address. Thus, it would show the EHLO address in Received:
4037 header lines when this was not necessary.
4039 PH/27 Added the % operator to ${eval:}.
4041 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4042 it should be ignoring failures (because with -C, for example, it has lost
4043 privilege). It wasn't ignoring creation failures other than "already
4046 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4049 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4050 because an input line was too long, either on its own, or by virtue of
4051 too many continuations, the temporary file was not being removed, and the
4052 return code was incorrect.
4054 PH/31 Missing "BOOL" in function definition in filtertest.c.
4056 PH/32 Applied Sieve patches from the maintainer.
4058 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4060 PH/33 Added "verify = not_blind".
4062 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4063 Local/Makefile (with some defaults set). These are used in built scripts
4064 such as exicyclog, but they have never been used in the exim_install
4065 script (though there are many overriding facilities there). I have
4066 arranged that the exim_install script now takes note of these two
4069 PH/35 Installed configuration files for Dragonfly.
4071 PH/36 When a locally submitted message by a trusted user did not contain a
4072 From: header, and the sender address was obtained from -f or from an SMTP
4073 MAIL command, and the trusted user did not use -F to supply a sender
4074 name, $originator_name was incorrectly used when constructing a From:
4075 header. Furthermore, $originator_name was used for submission mode
4076 messages from external hosts without From: headers in a similar way,
4077 which is clearly wrong.
4079 PH/37 Added control=suppress_local_fixups.
4081 PH/38 When log_selector = +received_sender was set, and the addition of the
4082 sender made the log line's construction buffer exactly full, or one byte
4083 less than full, an overflow happened when the terminating "\n" was
4086 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4087 when the result of a list match is failure because a DNS lookup failed.
4089 PH/40 RM_COMMAND is now used in the building process.
4091 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4092 the "build-* directories that it finds.
4094 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4095 address in a domain literal was a prefix of an interface address.
4097 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4098 when verifying a sender address, unless rewrite_headers is false.
4100 PH/44 Wrote a long comment about why errors_to addresses are verified as
4101 recipients, not senders.
4103 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4104 the ratelimit ACL was added.
4106 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4108 PH/46 Added extra information about PostgreSQL errors to the error string.
4110 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4111 avoiding the need to install special zones in a real server. This is
4112 backwards compatible; if it can't find the fake resolver, it drops back.
4113 Thus, both old and new test suites can be run.
4115 TF/02 Added util/ratelimit.pl
4117 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4118 clock is set back in time.
4120 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4121 Candler <B.Candler@pobox.com>.
4123 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4124 OK for addresses that are the result of redirections.
4126 PH/48 A number of further additions for the benefit of the new test suite,
4127 including a fake gethostbyname() that interfaces to the fake DNS resolver
4130 TF/06 The fix for widen_domains has also been applied to qualify_single and
4131 search_parents which are the other dnslookup options that can cause
4134 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4137 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4139 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4140 service" if it encountered an MX record with an empty target hostname.
4141 The message is now "an MX or SRV record indicated no SMTP service".
4143 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4144 if the DNS of the sending site is misconfigured. This is quite a
4145 common situation. This change restores the behaviour of treating a
4146 helo verification defer as a failure.
4148 PH/51 If self=fail was set on a router, the bounce message did not include the
4149 actual error message.
4155 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4157 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4158 command itself ended in error, the underlying message about the transport
4159 filter timeout was being overwritten with the pipe command error. Now the
4160 underlying error message should be appended to the second error message.
4162 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4164 PH/02 Reduce the amount of output that "make" produces by default. Full output
4165 can still be requested.
4167 PH/03 The warning log line about a condition test deferring for a "warn" verb
4168 was being output only once per connection, rather than after each
4169 occurrence (because it was using the same function as for successful
4170 "warn" verbs). This seems wrong, so I have changed it.
4172 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4173 it should not have, which might have caused a crash in the right
4174 circumstances, but probably never did.
4176 PH/04 Installed a modified version of Tony Finch's patch to make submission
4177 mode fix the return path as well as the Sender: header line, and to
4178 add a /name= option so that you can make the user's friendly name appear
4181 TF/03 Added the control = fakedefer ACL modifier.
4183 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4184 Mark Lowes for thorough testing.
4186 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4188 TK/03 Merged latest SRS patch from Miles Wilton.
4190 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4191 with the definition in sysexits.h (which is #included earlier).
4192 Fortunately, Exim does not actually use EX_OK. The code used to try to
4193 preserve the sysexits.h value, by assuming that macro definitions were
4194 scanned for macro replacements. I have been disabused of this notion,
4195 so now the code just undefines EX_OK before #including unistd.h.
4197 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4198 in the smtp transport. When a block could not be written in a single
4199 write() function, the timeout was being re-applied to each part-write.
4200 This seems wrong - if the receiver was accepting one byte at a time it
4201 would take for ever. The timeout is now adjusted when this happens. It
4202 doesn't have to be particularly precise.
4204 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4205 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4207 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4208 without a domain if the check to <postmaster@domain> fails.
4210 SC/01 Eximstats: added -xls and the ability to specify output files
4211 (patch written by Frank Heydlauf).
4213 SC/02 Eximstats: use FileHandles for outputting results.
4215 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4217 SC/04 Eximstats: fixed display of large numbers with -nvr option
4219 SC/05 Eximstats: fixed merging of reports with empty tables.
4221 SC/06 Eximstats: added the -include_original_destination flag
4223 SC/07 Eximstats: removed tabs and trailing whitespace.
4225 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4227 TK/06 MBOX spool code: Add real "From " MBOX separator line
4228 so the .eml file is really in mbox format (even though
4229 most programs do not really care). Patch from Alex Miller.
4231 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4232 The latter is generated from $received_to and is only set if the
4233 message has one envelope recipient. SA can use these headers,
4234 obviously out-of-the-box. Patch from Alex Miller.
4236 PH/08 The ${def test on a variable was returning false if the variable's
4237 value was "0", contrary to what the specification has always said!
4238 The result should be true unless the variable is empty.
4240 PH/09 The syntax error of a character other than { following "${if
4241 def:variable_name" (after optional whitespace) was not being diagnosed.
4242 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4243 accidental colon was present, for example, could give incorrect results.
4245 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4246 result is used (not including appendfile, where other changes are about
4249 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4250 This involved changing a lot of size variables from int to off_t. It
4251 should work with maildirs and everything.
4253 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4254 spamd dying while we are connected to it.
4256 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4259 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4260 function for BDB 4.3.
4262 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4264 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4265 the sources for PCRE have been re-arranged and more files are now
4268 PH/15 The code I had for printing potentially long long variables in PH/11
4269 above was not the best (it lost precision). The length of off_t variables
4270 is now inspected at build time, and an appropriate printing format (%ld
4271 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4272 to be "long long int" or "long int". This is needed for the internal
4273 formatting function string_vformat().
4275 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4276 the configuration file to be ":syslog", then the script "guesses" where
4277 the logs files are, rather than using the compiled in default. In our
4278 case the guess is not the same as the compiled default, so the script
4279 suddenly stopped working when I started to use syslog. The patch checks
4280 to see if log_file_path is "". If so, it attempts to read it from exim
4281 with no configuration file to get the compiled in version, before it
4282 falls back to the previous guessing code."
4284 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4285 implementing BATV in an Exim configuration. See NewStuff for the gory
4288 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4289 Makefile that are specific to HP-UX.
4291 PH/18 If the "use_postmaster" option was set for a recipient callout together
4292 with the "random" option, the postmaster address was used as the MAIL
4293 FROM address for the random test, but not for the subsequent recipient
4294 test. It is now used for both.
4296 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4297 patch removes a few documentation additions to RFC 3028, because the
4298 latest draft now contains them. It adds the new en;ascii-case comparator
4299 and a new error check for 8bit text in MIME parts. Comparator and
4300 require names are now matched exactly. I enabled the subaddress
4301 extension, but it is not well tested yet (read: it works for me)."
4303 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4304 rework some of the code of TK/09 above to avoid the hardwired use of
4305 "%lld" and "long long". Replaced the call to snprintf() with a call to
4308 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4309 records point to non-existent hosts", "retry timeout exceeded", and
4310 "retry time not reached for any host after a long failure period".
4312 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4313 experimental DomainKeys support:
4315 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4316 (2) On an error such as an illegally used "control", the wrong name for
4317 the control was given.
4319 These problems did NOT occur unless DomainKeys support was compiled.
4321 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4323 PH/24 Added ${if match_ip condition.
4325 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4326 databases so that it will be absolutely obvious if a crash occurs in the
4327 DB library. This is a regular occurrence (often caused by mis-matched
4330 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4331 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4332 picked up on a user's system that detects such things. There doesn't seem
4333 to be a gcc warning option for this - only an attribute that has to be
4334 put on the function's prototype. It seems that in Fedora Core 4 they have
4335 set this on a number of new functions. No doubt there will be more in due
4338 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4339 specify a transport. However, if an address that was verified by such a
4340 router was the subject of a callout, Exim crashed because it tried to
4341 read the rcpt_include_affixes from the non-existent transport. Now it
4342 just assumes that the setting of that option is false. This bug was
4343 introduced by 4.51/PH/31.
4345 PH/28 Changed -d+all to exclude +memory, because that information is very
4346 rarely of interest, but it makes the output a lot bigger. People tend to
4347 do -d+all out of habit.
4349 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4350 code in os-type was giving problems when libc.so lives in lib64, like on
4353 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4354 aren't the modern standard, and it seems that some systems' include files
4355 don't always have them. Exim was already checking for some of the newer
4356 ones like T_AAAA, and defining it itself. I've added checks for all the
4357 record types that Exim uses.
4359 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4360 not automatically generating a new one, as it is supposed to. This
4361 prevented TLS from working. If the file did exist, but contained invalid
4362 data, a new version was generated, as expected. It was only the case of a
4363 non-existent file that was broken.
4365 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4366 with a change in libdomainkeys > 0.64.
4368 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4369 from DNS. If the selector record carries the flag, it now has
4370 precedence over the domain-wide flag.
4372 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4374 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4375 the use of a port name that isn't defined in /etc/services) occurred, the
4376 message was deferred as in a normal delivery, and thus remained on the
4377 spool, instead of being failed because of the mua_wrapper setting. This
4378 is now fixed, and I tidied up some of the mua_wrapper messages at the
4381 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4382 the messages in a hash of arrays rather than using individual hashes.
4383 This is a bit cleaner and results in dramatic memory savings, albeit
4384 at a slight CPU cost.
4386 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4387 as requested by Marc Sherman.
4389 SC/10 Eximstats: added histograms for user specified patterns as requested
4392 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4394 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4395 fopen() in the content-scanning modules that did not already have it.
4401 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4402 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4404 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4406 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4408 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4409 to test Sieve filters that use "vacation".
4411 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4412 that changes the way the GnuTLS parameters are stored in the cache file.
4413 The new format can be generated externally. For backward compatibility,
4414 if the data in the cache doesn't make sense, Exim assumes it has read an
4415 old-format file, and it generates new data and writes a new file. This
4416 means that you can't go back to an older release without removing the
4419 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4420 work if there are any delivery delays because "one_time" forces the
4421 parent to be marked "delivered", so its unseen clone is never tried
4422 again. For this reason, Exim now forbids the simultaneous setting of
4425 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4426 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4427 entry if you want to know the details. The fix, however, neglected to
4428 consider the case when local delivery batching is involved. The test for
4429 "previously delivered" was not happening when checking to see if an
4430 address could be batched with a previous (undelivered) one; under
4431 certain circumstances this could lead to multiple deliveries to the same
4434 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4435 in its include files, and this causes problems building Exim.
4437 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4438 header_syntax) but Exim was just ignoring anything given after a slash.
4439 In particular, this caused confusion with an attempt to use "verify =
4440 reverse_host_lookup/defer_ok". An error is now given when options are
4441 supplied for verify items that do not have them. (Maybe reverse_host_
4442 lookup should have a defer_ok option, but that's a different point.)
4444 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4445 defined by RFC 821) to 2048, because there were problems with some AUTH
4446 commands, and RFC 1869 says the size should be increased for extended
4447 SMTP commands that take arguments.
4449 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4452 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4453 "unknown" error; now it says that the functionality isn't in the binary.
4455 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4456 an address' error message when a string expansion fails (syntax or
4457 whatever). Otherwise the password may appear in the log. Following change
4458 PH/42 below, there is no longer a chance of it appearing in a bounce
4461 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4463 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4464 of its addresses was ever tried. (Bugzilla bug #2.)
4466 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4467 the result incorrectly in the debug output. (It correctly added a newline
4468 to what was transported.)
4470 TF/01 Added $received_time.
4472 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4473 commented out examples of how to interface to a virus scanner and to
4474 SpamAssassin. Also added commented examples of av_scanner and
4475 spamd_address settings.
4477 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4478 and controls are allowed in which ACLs. There were a couple of minor
4479 errors. Some of the entries in the conditions table (which is a table of
4480 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4481 negation of where the condition IS allowed.
4483 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4485 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4486 header file does not have a version number, so I've had to invent a new
4487 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4488 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4489 radiusclient), but it was contributed by a Radius user.
4491 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4492 files or directories, for OpenSSL.
4494 PH/19 When an Exim process that is running as root has to create an Exim log
4495 file, it does so in a subprocess that runs as exim:exim so as to get the
4496 ownership right at creation (otherwise, other Exim processes might see
4497 the file with the wrong ownership). There was no test for failure of this
4498 fork() call, which would lead to the process getting stuck as it waited
4499 for a non-existent subprocess. Forks do occasionally fail when resources
4500 run out. I reviewed all the other calls to fork(); they all seem to check
4503 PH/20 When checking for unexpected SMTP input at connect time (before writing
4504 the banner), Exim was not dealing correctly with a non-positive return
4505 from the read() function. If the client had disconnected by this time,
4506 the result was a log entry for a synchronization error with an empty
4507 string after "input=" when read() returned zero. If read() returned -1
4508 (an event I could not check), uninitialized data bytes were printed.
4509 There were reports of junk text (parts of files, etc) appearing after
4512 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4514 PH/22 Added support for macro redefinition, and (re)definition in between
4515 driver and ACL definitions.
4517 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4518 forgetting to use the resulting value; it was using the unexpanded value.
4520 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4521 hadn't been configured. The fix is from Juergen Kreileder, who
4522 understands it better than I do:
4524 "Here's what I see happening with three configured cyrus_sasl
4525 authenticators configured (plain, login, cram-md5):
4527 On startup auth_cyrus_sasl_init() gets called for each of these.
4528 This means three calls to sasl_listmech() without a specified mech_list.
4529 => SASL tests which mechs of all available mechs actually work
4530 => three warnings about OTP not working
4531 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4533 With the patch, sasl_listmech() also gets called three times. But now
4534 SASL's mech_list option is set to the server_mech specified in the the
4535 authenticator. Or in other words, the answer from sasl_listmech()
4536 gets limited to just the mech you're testing for (which is different
4538 => the return list contains just 'plain' or 'login', 'cram-md5' or
4539 nothing depending on the value of ob->server_mech.
4541 I've just tested the patch: Authentication still works fine,
4542 unavailable mechs specified in the exim configuration are still
4543 caught, and the auth.log warnings about OTP are gone."
4545 PH/25 When debugging is enabled, the contents of the command line are added
4546 to the debugging output, even when log_selector=+arguments is not
4549 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4550 answer is "GNU", and only if the return is "GNU/something" is the answer
4553 PH/27 $acl_verify_message is now set immediately after the failure of a
4554 verification in an ACL, and so is available in subsequent modifiers. In
4555 particular, the message can be preserved by coding like this:
4557 warn !verify = sender
4558 set acl_m0 = $acl_verify_message
4560 Previously, $acl_verify_message was set only while expanding "message"
4561 and "log_message" when a very denied access.
4563 PH/28 Modified OS/os.c-Linux with
4565 -#ifndef OS_LOAD_AVERAGE
4566 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4568 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4569 nomenclature these days.)
4571 PH/29 Installed patch from the Sieve maintainer that adds the options
4572 sieve_useraddress and sieve_subaddress to the redirect router.
4574 PH/30 In these circumstances:
4575 . Two addresses routed to the same list of hosts;
4576 . First host does not offer TLS;
4577 . First host accepts first address;
4578 . First host gives temporary error to second address;
4579 . Second host offers TLS and a TLS session is established;
4580 . Second host accepts second address.
4581 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4582 and peerdn, if requested) that were in fact used only for the second
4585 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4586 attention to any local part prefix or suffix that was matched by the
4587 router that accepted the address. It now behaves in the same way as it
4588 does for delivery: the affixes are removed from the local part unless
4589 rcpt_include_affixes is set on the transport.
4591 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4592 timeout during the DATA phase of an incoming message.
4594 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4595 applied a patch sanctioned by the Sieve maintainer.
4597 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4598 the uid or gid is negative. A case of a negative gid caused this to be
4599 noticed. The fix allows for either to be negative.
4601 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4602 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4605 PH/36 The change PH/12 above was broken. Fixed it.
4607 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4608 the grounds that routing the same address twice would always produce the
4609 same answer. This might have been true once, but it is certainly no
4610 longer true now. Routing a child address may depend on the previous
4611 routing that produced that child. Some complicated redirection strategies
4612 went wrong when messages had multiple recipients, and made Exim's
4613 behaviour dependent on the order in which the addresses were given.
4615 I have moved the duplicate checking until after the routing is complete.
4616 Exim scans the addresses that are assigned to local and remote
4617 transports, and removes any duplicates. This means that more work will be
4618 done, as duplicates will always all be routed, but duplicates are
4619 presumably rare, so I don't expect this is of any significance.
4621 For deliveries to pipes, files, and autoreplies, the duplicate checking
4622 still happens during the routing process, since they are not going to be
4625 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4626 It corrects a timeout issue with spamd. This is Ian's comment: "The
4627 background is that sometimes spamd either never reads data from a
4628 connection it has accepted, or it never writes response data. The exiscan
4629 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4630 blindly assumes that writes won't block so it may never time out."
4632 PH/39 Allow G after quota size as well as K and M.
4634 PH/40 The value set for $authenticated_id in an authenticator may not contain
4635 binary zeroes or newlines because the value is written to log lines and
4636 to spool files. There was no check on this. Now the value is run through
4637 the string_printing() function so that such characters are converted to
4638 printable escape sequences.
4640 PH/41 $message_linecount is a new variable that contains the total number of
4641 lines in the message. Compare $body_linecount, which is the count for the
4644 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4645 bounce and delay warning messages, except in certain special cases, which
4648 (a) An SMTP error message from a remote host;
4649 (b) A message specified in a :fail: redirection;
4650 (c) A message specified in a "fail" command in a system filter;
4651 (d) A message specified in a FAIL return from the queryprogram router;
4652 (e) A message specified by the cannot_route_message router option.
4654 In these cases only, Exim does include the error details in bounce and
4655 warning messages. There are also a few cases where bland messages such
4656 as "unrouteable address" or "local delivery error" are given.
4658 PH/43 $value is now also set for the "else" part of a ${run expansion.
4660 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4661 being worked on, but at least Exim now implements the latest version to
4664 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4665 process to complete was treated as a delivery failure, a timeout while
4666 writing the message to the pipe was logged, but erroneously treated as a
4667 successful delivery. Such timeouts include transport filter timeouts. For
4668 consistency with the overall process timeout, these timeouts are now
4669 treated as errors, giving rise to delivery failures by default. However,
4670 there is now a new Boolean option for the pipe transport called
4671 timeout_defer, which, if set TRUE, converts the failures into defers for
4672 both kinds of timeout. A transport filter timeout is now identified in
4675 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4676 systems where "make" and "gmake" are different, calling "gmake" at top
4677 level broke things. I've arranged for the value of $(MAKE) to be passed
4678 from the Makefile to this script so that it can call the same version of
4682 A note about Exim versions 4.44 and 4.50
4683 ----------------------------------------
4685 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4686 changes of various kinds. As a consequence, a big documentation update was
4687 needed. This delayed the release for rather longer than seemed good, especially
4688 in the light of a couple of (minor) security issues. Therefore, the changes
4689 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4690 release. So 4.44 and 4.50 are in effect two different branches that both start
4693 I have left the 4.50 change log unchanged; it contains all the changes since
4694 4.43. The change log for 4.44 is below; many of its items are identical to
4695 those for 4.50. This seems to be the most sensible way to preserve the
4696 historical information.
4702 1. Minor wording change to the doc/README.SIEVE file.
4704 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4705 computation of the current number of files was incorrect.
4707 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4708 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4711 4. Give more explanation in the error message when the command for a transport
4712 filter fails to execute.
4714 5. There are several places where Exim runs a non-Exim command in a
4715 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4716 was being done only for the command run by the queryprogram router. It is
4717 now done for all such subprocesses. The other cases are: ${run, transport
4718 filters, and the commands run by the lmtp and pipe transports.
4720 6. Added CONFIGURE_GROUP build-time option.
4722 7. Some older OS have a limit of 256 on the maximum number of file
4723 descriptors. Exim was using setrlimit() to set 1000 as a large value
4724 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4725 systems. I've change it so that if it can't get 1000, it tries for 256.
4727 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4728 was an oversight, and furthermore, ever since the addition of extra
4729 controls (e.g. 4.43/32), the checks on when to allow different forms of
4730 "control" were broken. There should now be diagnostics for all cases when a
4731 control that does not make sense is encountered.
4733 9. Added the /retain_sender option to "control=submission".
4735 10. $recipients is now available in the predata ACL (oversight).
4737 11. Tidy the search cache before the fork to do a delivery from a message
4738 received from the command line. Otherwise the child will trigger a lookup
4739 failure and thereby defer the delivery if it tries to use (for example) a
4740 cached ldap connection that the parent has called unbind on.
4742 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4743 of $address_data from the recipient verification was clobbered by the
4744 sender verification.
4746 13. The value of address_data from a sender verification is now available in
4747 $sender_address_data in subsequent conditions in the ACL statement.
4749 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4751 15. Added a new option "connect=<time>" to callout options, to set a different
4754 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4755 was its contents. (It was OK if the option was not defined at all.)
4757 17. A "Completed" log line is now written for messages that are removed from
4758 the spool by the -Mrm option.
4760 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4761 information about exactly what failed.
4763 19. Added -dd to debug only the daemon process.
4765 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4766 handles timeouts, both on the server side and network timeouts. Renamed the
4767 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4769 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4770 It is now set to "smtps".
4772 22. $host_address is now set to the target address during the checking of
4773 ignore_target_hosts.
4775 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4776 being passed; this would have caused $sender_host_name to have been used if
4777 matching the list had actually called for a host name (not very likely,
4778 since this list is usually IP addresses). A host name is now passed as
4781 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4782 code that specifies a non-restarting handler (typically sigaction() in
4783 modern systems) in an attempt to fix a rare and obscure crash bug.
4785 25. Narrowed the window for a race in the daemon that could cause it to ignore
4786 SIGCHLD signals. This is not a major problem, because they are used only to
4787 wake it up if nothing else does.
4789 26. A malformed maildirsize file could cause Exim to calculate negative values
4790 for the mailbox size or file count. Odd effects could occur as a result.
4791 The maildirsize information is now recalculated if the size or filecount
4794 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4795 support for a long time. Removed HAVE_SYS_VFS_H.
4797 28. Installed the latest version of exipick from John Jetmore.
4799 29. In an address list, if the pattern was not a regular expression, an empty
4800 subject address (from a bounce message) matched only if the pattern was an
4801 empty string. Non-empty patterns were not even tested. This was the wrong
4802 because it is perfectly reasonable to use an empty address as part of a
4803 database query. An empty address is now tested by patterns that are
4804 lookups. However, all the other forms of pattern expect the subject to
4805 contain a local part and a domain, and therefore, for them, an empty
4806 address still always fails if the pattern is not itself empty.
4808 30. Exim went into a mad DNS loop when attempting to do a callout where the
4809 host was specified on an smtp transport, and looking it up yielded more
4810 than one IP address.
4812 31. Re-factored the code for checking spool and log partition space into a
4813 function that finds that data and another that does the check. The former
4814 is then used to implement four new variables: $spool_space, $log_space,
4815 $spool_inodes, and $log_inodes.
4817 32. The RFC2047 encoding function was originally intended for short strings
4818 such as real names; it was not keeping to the 75-character limit for
4819 encoded words that the RFC imposes. It now respects the limit, and
4820 generates multiple encoded words if necessary. To be on the safe side, I
4821 have increased the buffer size for the ${rfc2047: expansion operator from
4824 33. It is now permitted to omit both strings after an "if" condition; if the
4825 condition is true, the result is "true". As before, when the second string
4826 is omitted, a false condition yields an empty string. This makes it less
4827 cumbersome to write custom ACL and router conditions.
4829 34. Failure to deliver a bounce message always caused it to be frozen, even if
4830 there was an errors_to setting on the router. The errors_to setting is now
4833 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4834 canonical form (fully expanded) before being placed in
4835 $sender_host_address.
4837 36. The table in the code that translates DNS record types into text (T_A to
4838 "A" for instance) was missing entries for NS and CNAME. It is just possible
4839 that this could have caused confusion if both these types were looked up
4840 for the same domain, because the text type is used as part of Exim's
4841 per-process caching. But the chance of anyone hitting this buglet seems
4844 37. The dnsdb lookup has been extended in a number of ways.
4846 (1) There is a new type, "zns", which walks up the domain tree until it
4847 finds some nameserver records. It should be used with care.
4849 (2) There is a new type, "mxh", which is like "mx" except that it returns
4850 just the host names, not the priorities.
4852 (3) It is now possible to give a list of domains (or IP addresses) to be
4853 looked up. The behaviour when one of the lookups defers can be
4854 controlled by a keyword.
4856 (4) It is now possible to specify the separator character for use when
4857 multiple records are returned.
4859 38. The dnslists ACL condition has been extended: it is now possible to supply
4860 a list of IP addresses and/or domains to be looked up in a particular DNS
4863 39. Added log_selector=+queue_time_overall.
4865 40. When running the queue in the test harness, wait just a tad after forking a
4866 delivery process, to get repeatability of debugging output.
4868 41. Include certificate and key file names in error message when GnuTLS fails
4869 to set them up, because the GnuTLS error message doesn't include the name
4870 of the failing file when there is a problem reading it.
4872 42. Allow both -bf and -bF in the same test run.
4874 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4876 44. The "Exiscan patch" is now merged into the mainline Exim source.
4878 45. Sometimes the final signoff response after QUIT could fail to get
4879 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4880 < 0 before doing a fflush(). This bug looks as though it goes back to the
4881 introduction of TLS in release 3.20, but "sometimes" must have been rare
4882 because the tests only now provoked it.
4884 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4885 (this can affect the format of dates).
4887 47. exim_tidydb, when checking for the continued existence of a message for
4888 which it has found a message-specific retry record, was not finding
4889 messages that were in split spool directories. Consequently, it was
4890 deleting retry records that should have stayed in existence.
4892 48. Steve fixed some bugs in eximstats.
4894 49. The SPA authentication driver was not abandoning authentication and moving
4895 on to the next authenticator when an expansion was forced to fail,
4896 contradicting the general specification for all authenticators. Instead it
4897 was generating a temporary error. It now behaves as specified.
4899 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4900 (the order specifies the preference for clients). The order is now AES256,
4901 AES128, 3DES, ARCFOUR128.
4903 51. Small patch to Sieve code - explicitly set From: when generating an
4906 52. Exim crashed if a remote delivery caused a very long error message to be
4907 recorded - for instance if somebody sent an entire SpamAssassin report back
4908 as a large number of 550 error lines. This bug was coincidentally fixed by
4909 increasing the size of one of Exim's internal buffers (big_buffer) that
4910 happened as part of the Exiscan merge. However, to be on the safe side, I
4911 have made the code more robust (and fixed the comments that describe what
4914 53. Now that there can be additional text after "Completed" in log lines (if
4915 the queue_time_overall log selector is set), a one-byte patch to exigrep
4916 was needed to allow it to recognize "Completed" as not the last thing in
4919 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4920 patch that reportedly fixes this has been added. I am not expert enough to
4921 create a test for it. This is what the patch creator wrote:
4923 "I found a little strange behaviour of ldap code when working with
4924 Windows 2003 AD Domain, where users was placed in more than one
4925 Organization Units. When I tried to give exim partial DN, the exit code
4926 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4927 But simultaneously result of request was absolutely normal ldap result,
4928 so I produce this patch..."
4930 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4931 REFERENCE, so I have modified the code to exclude the patch when that macro
4934 55. Some experimental protocols are using DNS PTR records for new purposes. The
4935 keys for these records are domain names, not reversed IP addresses. The
4936 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4937 leaves it alone. Component reversal etc. now happens only for IP addresses.
4940 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4942 57. Double the size of the debug message buffer (to 2048) so that more of very
4943 long debug lines gets shown.
4945 58. The exicyclog utility now does better if the number of log files to keep
4946 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4948 59. Two changes related to the smtp_active_hostname option:
4950 (1) $smtp_active_hostname is now available as a variable.
4951 (2) The default for smtp_banner uses $smtp_active_hostname instead
4952 of $primary_hostname.
4954 60. The host_aton() function is supposed to be passed a string that is known
4955 to be a valid IP address. However, in the case of IPv6 addresses, it was
4956 not checking this. This is a hostage to fortune. Exim now panics and dies
4957 if the condition is not met. A case was found where this could be provoked
4958 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4959 components; fortuitously, this particular loophole had already been fixed
4960 by change 4.50/55 above.
4962 If there are any other similar loopholes, the new check in host_aton()
4963 itself should stop them being exploited. The report I received stated that
4964 data on the command line could provoke the exploit when Exim was running as
4965 exim, but did not say which command line option was involved. All I could
4966 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4967 running as the user.
4970 61. There was a buffer overflow vulnerability in the SPA authentication code
4971 (which came originally from the Samba project). I have added a test to the
4972 spa_base64_to_bits() function which I hope fixes it.
4975 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4976 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4978 63. The daemon start-up calls getloadavg() while still root for those OS that
4979 need the first call to be done as root, but it missed one case: when
4980 deliver_queue_load_max is set with deliver_drop_privilege. This is
4981 necessary for the benefit of the queue runner, because there is no re-exec
4982 when deliver_drop_privilege is set.
4984 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4985 This has been fixed.
4987 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4988 was in use, was not putting the data itself into the right store pool;
4989 consequently, it could be overwritten for a subsequent message in the same
4990 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4993 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4995 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4996 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4997 the function were treating the return as a boolean value, which happened to
4998 work because 0=false and not-0=true, but is not correct code.
5000 68. The host_aton() function was not handling scoped IPv6 addresses (those
5001 with, for example, "%eth0" on the end) correctly.
5003 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5004 negated items (that is, ~something) in unsigned ints. Some compilers
5005 apparently mutter when there is no cast.
5007 70. If an address verification called from an ACL failed, and did not produce a
5008 user-specific message (i.e. there was only a "system" message), nothing was
5009 put in $acl_verify_message. In this situation, it now puts the system
5012 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5013 session; change 4.31/43 added the unwanted input to the log line - except
5014 that it did not do this in the start of session case. It now does.
5016 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5017 This is wrong and can cause the other end to generate a synchronization
5018 error if it is another Exim or anything else that does the synchronization
5019 check. A QUIT command is no longer sent after a timeout.
5021 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5022 during host lookups.
5024 74. The defer_ok option of callout verification was not working if it was used
5025 when verifying addresses in header lines, that is, for this case:
5027 verify = header_sender/callout=defer_ok
5029 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5030 those file descriptors could be used for SMTP connections. If anything
5031 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5032 could be sent to the SMTP client, causing chaos. The daemon now opens
5033 stdin, stdout, and stderr to /dev/null when it puts itself into the
5036 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5037 log by default. The user can override this with suitable Perl magic.
5039 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5040 the log message when discard triggers, was not working for the DATA ACL or
5041 for the non-SMTP ACL.
5043 78. Error message wording change in sieve.c.
5045 79. If smtp_accept_max_per_host was set, the number of connections could be
5046 restricted to fewer than expected, because the daemon was trying to set up
5047 a new connection before checking whether the processes handling previous
5048 connections had finished. The check for completed processes is now done
5049 earlier. On busy systems, this bug wouldn't be noticed because something
5050 else would have woken the daemon, and it would have reaped the completed
5053 80. If a message was submitted locally by a user whose login name contained one
5054 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5055 It caused a spool format error. I have fixed the spool reading code. A
5056 related problem was that the "from" clause in the Received: line became
5057 illegal because of the space(s). It is now covered by ${quote_local_part.
5059 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5062 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5064 83. Patch from Sieve maintainer to fix unterminated string problem in
5065 "vacation" handling.
5067 84. Some minor changes to the Linux configuration files to help with other
5068 OS variants using glibc.
5070 85. One more patch for Sieve to update vacation handling to latest spec.
5073 ----------------------------------------------------
5074 See the note above about the 4.44 and 4.50 releases.
5075 ----------------------------------------------------
5081 1. Change 4.43/35 introduced a bug that caused file counts to be
5082 incorrectly computed when quota_filecount was set in an appendfile
5085 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5086 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5089 3. Give more explanation in the error message when the command for a transport
5090 filter fails to execute.
5092 4. There are several places where Exim runs a non-Exim command in a
5093 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5094 was being done only for the command run by the queryprogram router. It is
5095 now done for all such subprocesses. The other cases are: ${run, transport
5096 filters, and the commands run by the lmtp and pipe transports.
5098 5. Some older OS have a limit of 256 on the maximum number of file
5099 descriptors. Exim was using setrlimit() to set 1000 as a large value
5100 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5101 systems. I've change it so that if it can't get 1000, it tries for 256.
5103 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5104 was an oversight, and furthermore, ever since the addition of extra
5105 controls (e.g. 4.43/32), the checks on when to allow different forms of
5106 "control" were broken. There should now be diagnostics for all cases when a
5107 control that does not make sense is encountered.
5109 7. $recipients is now available in the predata ACL (oversight).
5111 8. Tidy the search cache before the fork to do a delivery from a message
5112 received from the command line. Otherwise the child will trigger a lookup
5113 failure and thereby defer the delivery if it tries to use (for example) a
5114 cached ldap connection that the parent has called unbind on.
5116 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5117 of $address_data from the recipient verification was clobbered by the
5118 sender verification.
5120 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5121 was its contents. (It was OK if the option was not defined at all.)
5123 11. A "Completed" log line is now written for messages that are removed from
5124 the spool by the -Mrm option.
5126 12. $host_address is now set to the target address during the checking of
5127 ignore_target_hosts.
5129 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5130 being passed; this would have caused $sender_host_name to have been used if
5131 matching the list had actually called for a host name (not very likely,
5132 since this list is usually IP addresses). A host name is now passed as
5135 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5136 code that specifies a non-restarting handler (typically sigaction() in
5137 modern systems) in an attempt to fix a rare and obscure crash bug.
5139 15. Narrowed the window for a race in the daemon that could cause it to ignore
5140 SIGCHLD signals. This is not a major problem, because they are used only to
5141 wake it up if nothing else does.
5143 16. A malformed maildirsize file could cause Exim to calculate negative values
5144 for the mailbox size or file count. Odd effects could occur as a result.
5145 The maildirsize information is now recalculated if the size or filecount
5148 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5149 support for a long time. Removed HAVE_SYS_VFS_H.
5151 18. Updated exipick to current release from John Jetmore.
5153 19. Allow an empty sender to be matched against a lookup in an address list.
5154 Previously the only cases considered were a regular expression, or an
5157 20. Exim went into a mad DNS lookup loop when doing a callout where the
5158 host was specified on the transport, if the DNS lookup yielded more than
5161 21. The RFC2047 encoding function was originally intended for short strings
5162 such as real names; it was not keeping to the 75-character limit for
5163 encoded words that the RFC imposes. It now respects the limit, and
5164 generates multiple encoded words if necessary. To be on the safe side, I
5165 have increased the buffer size for the ${rfc2047: expansion operator from
5168 22. Failure to deliver a bounce message always caused it to be frozen, even if
5169 there was an errors_to setting on the router. The errors_to setting is now
5172 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5173 canonical form (fully expanded) before being placed in
5174 $sender_host_address.
5176 24. Updated eximstats to version 1.33
5178 25. Include certificate and key file names in error message when GnuTLS fails
5179 to set them up, because the GnuTLS error message doesn't include the name
5180 of the failing file when there is a problem reading it.
5182 26. Expand error message when OpenSSL has problems setting up cert/key files.
5185 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5186 (this can affect the format of dates).
5188 28. exim_tidydb, when checking for the continued existence of a message for
5189 which it has found a message-specific retry record, was not finding
5190 messages that were in split spool directories. Consequently, it was
5191 deleting retry records that should have stayed in existence.
5193 29. eximstats updated to version 1.35
5194 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5195 1.35 - bugfix such that pie charts by volume are generated correctly
5197 30. The SPA authentication driver was not abandoning authentication and moving
5198 on to the next authenticator when an expansion was forced to fail,
5199 contradicting the general specification for all authenticators. Instead it
5200 was generating a temporary error. It now behaves as specified.
5202 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5203 (the order specifies the preference for clients). The order is now AES256,
5204 AES128, 3DES, ARCFOUR128.
5206 31. Small patch to Sieve code - explicitly set From: when generating an
5209 32. Exim crashed if a remote delivery caused a very long error message to be
5210 recorded - for instance if somebody sent an entire SpamAssassin report back
5211 as a large number of 550 error lines. This bug was coincidentally fixed by
5212 increasing the size of one of Exim's internal buffers (big_buffer) that
5213 happened as part of the Exiscan merge. However, to be on the safe side, I
5214 have made the code more robust (and fixed the comments that describe what
5217 33. Some experimental protocols are using DNS PTR records for new purposes. The
5218 keys for these records are domain names, not reversed IP addresses. The
5219 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5220 leaves it alone. Component reversal etc. now happens only for IP addresses.
5223 34. The host_aton() function is supposed to be passed a string that is known
5224 to be a valid IP address. However, in the case of IPv6 addresses, it was
5225 not checking this. This is a hostage to fortune. Exim now panics and dies
5226 if the condition is not met. A case was found where this could be provoked
5227 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5228 components; fortuitously, this particular loophole had already been fixed
5229 by change 4.50/55 or 4.44/33 above.
5231 If there are any other similar loopholes, the new check in host_aton()
5232 itself should stop them being exploited. The report I received stated that
5233 data on the command line could provoke the exploit when Exim was running as
5234 exim, but did not say which command line option was involved. All I could
5235 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5236 running as the user.
5239 35. There was a buffer overflow vulnerability in the SPA authentication code
5240 (which came originally from the Samba project). I have added a test to the
5241 spa_base64_to_bits() function which I hope fixes it.
5244 36. The daemon start-up calls getloadavg() while still root for those OS that
5245 need the first call to be done as root, but it missed one case: when
5246 deliver_queue_load_max is set with deliver_drop_privilege. This is
5247 necessary for the benefit of the queue runner, because there is no re-exec
5248 when deliver_drop_privilege is set.
5250 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5251 was in use, was not putting the data itself into the right store pool;
5252 consequently, it could be overwritten for a subsequent message in the same
5253 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5256 38. Sometimes the final signoff response after QUIT could fail to get
5257 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5258 < 0 before doing a fflush(). This bug looks as though it goes back to the
5259 introduction of TLS in release 3.20, but "sometimes" must have been rare
5260 because the tests only now provoked it.
5266 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5267 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5268 Now it returns an integer. A number of calls were still expecting a T/F
5269 return. Fortuitously, in all cases, the tests worked in OK situations,
5270 which is the norm. However, things would have gone wrong on any write
5271 failures on the smtp file descriptor. This function is used when sending
5272 messages over SMTP and also when doing verify callouts.
5274 2. When Exim is called to do synchronous delivery of a locally submitted
5275 message (the -odf or -odi options), it no longer closes stderr before doing
5278 3. Implemented the mua_wrapper option.
5280 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5282 5. Implemented the functions header_remove(), header_testname(),
5283 header_add_at_position(), and receive_remove_recipient(), and exported them
5286 6. If an ACL "warn" statement specified the addition of headers, Exim already
5287 inserted X-ACL-Warn: at the start if there was no header name. However, it
5288 was not making this test for the second and subsequent header lines if
5289 there were newlines in the string. This meant that an invalid header could
5290 be inserted if Exim was badly configured.
5292 7. Allow an ACL "warn" statement to add header lines at the start or after all
5293 the Received: headers, as well as at the end.
5295 8. Added the rcpt_4xx retry error code.
5297 9. Added postmaster_mailfrom=xxx to callout verification option.
5299 10. Added mailfrom=xxxx to the callout verification option, for verify=
5302 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5303 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5305 12. Inserted some casts to stop certain compilers warning when using pointer
5306 differences as field lengths or precisions in printf-type calls (mostly
5307 affecting debugging statements).
5309 13. Added optional readline() support for -be (dynamically loaded).
5311 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5312 same clock tick as a message's arrival, so that its received time was the
5313 same as the "first fail" time on the retry record, and that message
5314 remained on the queue past the ultimate address timeout, every queue runner
5315 would try a delivery (because it was past the ultimate address timeout) but
5316 after another failure, the ultimate address timeout, which should have then
5317 bounced the address, did not kick in. This was a "< instead of <=" error;
5318 in most cases the first failure would have been in the next clock tick
5319 after the received time, and all would be well.
5321 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5322 being recognized when the domain list was tested by the match_domain
5323 condition in an expansion string.
5325 16. Added the ${str2b64: operator.
5327 17. Exim was always calling setrlimit() to set a large limit for the number of
5328 processes, without checking whether the existing limit was already
5329 adequate. (It did check for the limit on file descriptors.) Furthermore,
5330 errors from getrlimit() and setrlimit() were being ignored. Now they are
5331 logged to the main and panic logs, but Exim does carry on, to try to do its
5332 job under whatever limits there are.
5334 18. Imported PCRE 5.0.
5336 19. Trivial typo in log message " temporarily refused connection" (the leading
5339 20. If the log selector return_path_on_delivery was set and an address was
5340 redirected to /dev/null, the delivery process crashed because it assumed
5341 that a return path would always be set for a "successful" delivery. In this
5342 case, the whole delivery is bypassed as an optimization, and therefore no
5345 21. Internal re-arrangement: the function for sending a challenge and reading
5346 a response while authentication was assuming a zero-terminated challenge
5347 string. It's now changed to take a pointer and a length, to allow for
5348 binary data in such strings.
5350 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5352 23. Exim was not respecting finduser_retries when seeking the login of the
5353 uid under which it was called; it was always trying 10 times. (The default
5354 setting of finduser_retries is zero.) Also, it was sleeping after the final
5355 failure, which is pointless.
5357 24. Implemented tls_on_connect_ports.
5359 25. Implemented acl_smtp_predata.
5361 26. If the domain in control=submission is set empty, Exim assumes that the
5362 authenticated id is a complete email address when it generates From: or
5363 Sender: header lines.
5365 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5366 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5367 chown and chgrp in /bin and hostname in /usr/bin.
5369 28. Exim was keeping the "process log" file open after each use, just as it
5370 does for the main log. This opens the possibility of it remaining open for
5371 long periods when the USR1 signal hits a daemon. Occasional processlog
5372 errors were reported, that could have been caused by this. Anyway, it seems
5373 much more sensible not to leave this file open at all, so that is what now
5376 29. The long-running daemon process does not normally write to the log once it
5377 has entered its main loop, and it closes the log before doing so. This is
5378 so that log files can straightforwardly be renamed and moved. However,
5379 there are a couple of unusual error situations where the daemon does write
5380 log entries, and I had neglected to close the log afterwards.
5382 30. The text of an SMTP error response that was received during a remote
5383 delivery was being truncated at 512 bytes. This is too short for some of
5384 the long messages that one sometimes sees. I've increased the limit to
5387 31. It is now possible to make retry rules that apply only when a message has a
5388 specific sender, in particular, an empty sender.
5390 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5391 it possible to be selective about when SMTP synchronization is enforced.
5393 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5395 32. Implemented hosts_connection_nolog.
5397 33. Added an ACL for QUIT.
5399 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5402 35. Added mailbox_size and mailbox_filecount to appendfile.
5404 36. Added control = no_multiline_responses to ACLs.
5406 37. There was a bug in the logic of the code that waits for the clock to tick
5407 in the case where the clock went backwards by a substantial amount such
5408 that the microsecond fraction of "now" was more than the microsecond
5409 fraction of "then" (but the whole seconds number was less).
5411 38. Added support for the libradius Radius client library this is found on
5412 FreeBSD (previously only the radiusclient library was supported).
5418 1. When certain lookups returned multiple values in the form name=value, the
5419 quoting of the values was not always being done properly. Specifically:
5420 (a) If the value started with a double quote, but contained no whitespace,
5422 (b) If the value contained whitespace other than a space character (i.e.
5423 tabs or newlines or carriage returns) it was not quoted.
5424 This fix has been applied to the mysql and pgsql lookups by writing a
5425 separate quoting function and calling it from the lookup code. The fix
5426 should probably also be applied to nisplus, ibase and oracle lookups, but
5427 since I cannot test any of those, I have not disturbed their existing code.
5429 2. A hit in the callout cache for a specific address caused a log line with no
5430 reason for rejecting RCPT. Now it says "Previous (cached) callout
5431 verification failure".
5433 3. There was an off-by-one bug in the queryprogram router. An over-long
5434 return line was truncated at 256 instead of 255 characters, thereby
5435 overflowing its buffer with the terminating zero. As well as fixing this, I
5436 have increased the buffer size to 1024 (and made a note to document this).
5438 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5439 when Exim is waiting for an SMTP response from a remote server, Exim
5440 restarts its select() call on the socket, thereby resetting its timeout.
5441 This is not a problem when such interrupts are rare. Somebody set up a cron
5442 job to run exiwhat every 2 minutes, which is less than the normal select()
5443 timeout (5 or 10 minutes). This meant that the select() timeout never
5444 kicked in because it was always reset. I have fixed this by comparing the
5445 time when an interrupt arrives with the time at the start of the first call
5446 to select(). If more time than the timeout has elapsed, the interrupt is
5447 treated as a timeout.
5449 5. Some internal re-factoring in preparation for the addition of Sieve
5450 extensions (by MH). In particular, the "personal" test is moved to a
5451 separate function, and given an option for scanning Cc: and Bcc: (which is
5452 not set for Exim filters).
5454 6. When Exim created an email address using the login of the caller as the
5455 local part (e.g. when creating a From: or Sender: header line), it was not
5456 quoting the local part when it contained special characters such as @.
5458 7. Installed new OpenBSD configuration files.
5460 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5461 try to make them clearer.
5463 9. Callout options, other than the timeout value, were being ignored when
5464 verifying sender addresses in header lines. For example, when using
5466 verify = header_sender/callout=no_cache
5468 the cache was (incorrectly) being used.
5470 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5471 only the Cygwin environment.
5473 11. When return_path_on_delivery was set as a log selector, if different remote
5474 addresses in the same message used different return paths and parallel
5475 remote delivery occurred, the wrong values would sometimes be logged.
5476 (Whenever a remote delivery process finished, the return path value from
5477 the most recently started remote delivery process was logged.)
5479 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5480 lines when AUTH and/or TLS are in use. This is the "received protocol"
5481 field. Exim used to use "asmtp" for authenticated SMTP, without any
5482 indication (in the protocol name) for TLS use. Now it follows the RFC and
5483 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5484 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5485 names appear in log lines as well as in Received: header lines.
5487 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5488 extensions, and comparison tests, and to fix some bugs.
5490 14. Changes to the "personal" filter test:
5492 (1) The test was buggy in that it was just doing the equivalent of
5493 "contains" tests on header lines. For example, if a user's address was
5494 anne@some.where, the "personal" test would incorrectly be true for
5496 To: susanne@some.where
5498 This test is now done by extracting each address from the header in turn,
5499 and checking the entire address. Other tests that are part of "personal"
5500 are now done using regular expressions (for example, to check local parts
5501 of addresses in From: header lines).
5503 (2) The list of non-personal local parts in From: addresses has been
5504 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5505 taken from the Sieve specification recommendations.
5507 (3) If the message contains any header line starting with "List-" it is
5508 treated as non-personal.
5510 (4) The test for "circular" in the Subject: header line has been removed
5511 because it now seems ill-conceived.
5513 15. Minor typos in src/EDITME comments corrected.
5515 16. Installed latest exipick from John Jetmore.
5517 17. If headers_add on a router specified a text string that was too long for
5518 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5519 of string_sprintf() is now avoided.
5521 18. $message_body_size was not set (it was always zero) when running the DATA
5522 ACL and the local_scan() function.
5524 19. For the "mail" command in an Exim filter, no default was being set for
5525 the once_repeat time, causing a random time value to be used if "once" was
5526 specified. (If the value happened to be <= 0, no repeat happened.) The
5527 default is now 0s, meaning "never repeat". The "vacation" command was OK
5528 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5529 (I found it when inspecting the code).
5531 20. There is now an overall timeout for performing a callout verification. It
5532 defaults to 4 times the callout timeout, which applies to individual SMTP
5533 commands during the callout. The overall timeout applies when there is more
5534 than one host that can be tried. The timeout is checked before trying the
5535 next host. This prevents very long delays if there are a large number of
5536 hosts and all are timing out (e.g. when the network connections are timing
5537 out). The value of the overall timeout can be changed by specifying an
5538 additional sub-option for "callout", called "maxwait". For example:
5540 verify = sender/callout=5s,maxwait=20s
5542 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5543 to the end before writing, but this should make it even safer).
5545 22. Exim was forgetting that it had advertised PIPELINING for the second and
5546 subsequent messages on an SMTP connection. It was also not resetting its
5547 memory on STARTTLS and an internal HELO.
5549 23. When Exim logs an SMTP synchronization error within a session, it now
5550 records whether PIPELINING has been advertised or not.
5552 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5553 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5554 rather than long int.
5556 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5558 26. Added the never_mail option to autoreply.
5564 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5565 crash if the getsockname() call failed; this can happen if a connection is
5566 closed very soon after it is established. The problem was simply in the
5567 order in which certain operations were done, causing Exim to try to write
5568 to the SMTP stream before it had set up the file descriptor. The bug has
5569 been fixed by making things happen in the correct order.
5575 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5576 before the connection was closed, thus losing the rejection response.
5578 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5579 some early Solaris releases, but causes trouble in current releases where
5580 socklen_t is defined.
5582 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5585 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5588 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5589 that releasing the top of it at the end releases what was used for sub-
5590 expansions (unless the block got too big). However, discard this block if
5591 the first thing is a variable or header, so that we can use its block when
5592 it is dynamic (useful for very large $message_headers, for example).
5594 6. Lookups now cache *every* query, not just the most recent. A new, separate
5595 store pool is used for this. It can be recovered when all lookup caches are
5596 flushed. Lookups now release memory at the end of their result strings.
5597 This has involved some general refactoring of the lookup sources.
5599 7. Some code has been added to the store_xxx() functions to reduce the amount
5600 of flapping under certain conditions.
5602 8. log_incoming_interface used to affect only the <= reception log lines. Now
5603 it causes the local interface and port to be added to several more SMTP log
5604 lines, for example "SMTP connection from", and rejection lines.
5606 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5608 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5610 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5611 could be overwritten at the end of the current message (or the start of a
5612 new message if it was set in a HELO ACL). The value is now preserved for
5613 the duration of the SMTP connection.
5615 12. If a transport had a headers_rewrite setting, and a matching header line
5616 contained an unqualified address, that address was qualified, even if it
5617 did not match any rewriting rules. The underlying bug was that the values
5618 of the flags that permit the existence of unqualified sender and recipient
5619 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5620 non-local messages, and by -bnq for local messages) were not being
5621 preserved with the message after it was received.
5623 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5624 "next input=" as part of the text comprising the host identity instead of
5625 the correct text. The code was using the same buffer for two different
5626 strings. However, depending on which order the printing function evaluated
5627 its arguments, the bug did not always show up. Under Linux, for example, my
5628 test suite worked just fine.
5630 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5631 doesn't work with some older versions of Perl. It has been changed to "my",
5632 which in any case is probably the better facility to use.
5634 15. A really picky compiler found some instances of statements for creating
5635 error messages that either had too many or two few arguments for the format
5638 16. The size of the buffer for calls to the DNS resolver has been increased
5639 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5640 for addresses that have a lot of PTR records. This alleviates a problem; it
5641 does not fully solve it.
5643 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5644 buffer now truncates the list and logs the incident, which is the same
5645 action as happens when Exim is looking up a host name and its aliases.
5646 Previously in this situation something unpredictable would happen;
5647 sometimes it was "internal error: store_reset failed".
5649 18. If a server dropped the connection unexpectedly when an Exim client was
5650 using GnuTLS and trying to read a response, the client delivery process
5651 crashed while trying to generate an error log message.
5653 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5654 string, for example:
5656 warn message = H1: something\nH2: something
5658 the text was added as a single header line from Exim's point of view
5659 though it ended up OK in the delivered message. However, searching for the
5660 second and subsequent header lines using $h_h2: did not work. This has been
5661 fixed. Similarly, if a system filter added multiple headers in this way,
5662 the routers could not see them.
5664 20. Expanded the error message when iplsearch is called with an invalid key to
5665 suggest using net-iplsearch in a host list.
5667 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5668 ACLs are no longer actually imposed (and a message to that effect is
5671 22. If a "gecos" field in a passwd entry contained escaped characters, in
5672 particular, if it contained a \" sequence, Exim got it wrong when building
5673 a From: or a Sender: header from that name. A second bug also caused
5674 incorrect handling when an unquoted " was present following a character
5675 that needed quoting.
5677 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5678 was not being matched caselessly.
5680 24. Arranged for all hyphens in the exim.8 source to be escaped with
5683 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5684 properly. Recipient callout cache records were still being keyed to include
5685 the sender, even when use_sender was set false. This led to far more
5686 callouts that were necessary. The sender is no longer included in the key
5687 when use_sender is false.
5689 26. Added "control = submission" modifier to ACLs.
5691 27. Added the ${base62d: operator to decode base 62 numbers.
5693 28. dnsdb lookups can now access SRV records.
5695 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5696 the configuration file.
5698 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5699 (-v) mode. This makes the output for a verbose queue run more intelligible.
5701 31. Added a use_postmaster feature to recipient callouts.
5703 32. Added the $body_zerocount variable, containing the number of binary zero
5704 bytes in the message body.
5706 33. The time of last modification of the "new" subdirectory is now used as the
5707 "mailbox time last read" when there is a quota error for a maildir
5710 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5712 35. Added +ignore_unknown as a special item in host lists.
5714 36. Code for decoding IPv6 addresses in host lists is now included, even if
5715 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5716 address was recognized as an IP address, but was then not correctly decoded
5717 into binary, causing unexpected and incorrect effects when compared with
5724 1. Very minor rewording of debugging text in manualroute to say "list of
5725 hosts" instead of "hostlist".
5727 2. If verify=header_syntax was set, and a header line with an unqualified
5728 address (no domain) and a large number of spaces between the end of the
5729 name and the colon was received, the reception process suffered a buffer
5730 overflow, and (when I tested it) crashed. This was caused by some obsolete
5731 code that should have been removed. The fix is to remove it!
5733 3. When running in the test harness, delay a bit after writing a bounce
5734 message to get a bit more predictability in the log output.
5736 4. Added a call to search_tidyup() just before forking a reception process. In
5737 theory, someone could use a lookup in the expansion of smtp_accept_max_
5738 per_host which, without the tidyup, could leave open a database connection.
5740 5. Added the variables $recipient_data and $sender_data which get set from a
5741 lookup success in an ACL "recipients" or "senders" condition, or a router
5742 "senders" option, similar to $domain_data and $local_part_data.
5744 6. Moved the writing of debug_print from before to after the "senders" test
5747 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5748 problems for message scanning, either using a data ACL, or using
5749 local_scan() because the Received: header was not generated till after they
5750 were called (in order to set the time as the time of reception completion).
5751 I have revised the way this works. The header is now generated after the
5752 body is received, but before the ACL or local_scan() are called. After they
5753 are run, the timestamp in the header is updated.
5759 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5760 before starting a queue runner without re-exec. This happened only when
5761 deliver_drop_privilege was set or when the Exim user was set to root. The
5762 effect of the bug was that timeouts during subsequent deliveries caused
5763 crashes instead of being properly handled. The handler is now left at its
5764 default (and expected) setting.
5766 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5767 message, again when deliver_drop_privilege is set or Exim is run as root.
5768 The bug described in (1) was not present in this case, but the tidying up
5769 of the other signals was missing. I have made the two cases consistent.
5771 3. The ignore_target_hosts setting on a manualroute router was being ignored
5772 for hosts that were looked up using the /MX notation.
5774 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5777 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5778 operated on the sender address. After changing the $sender_address to <>
5779 for the sender address verify, Exim was re-instated it as the original
5780 (before rewriting) address, but remembering that it had rewritten it, so it
5781 wasn't rewriting it again. This bug also had the effect of breaking the
5782 sender address verification caching when the sender address was rewritten.
5784 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5785 This has been changed so that if the ip literal address matches
5786 ignore_target_hosts, the router declines.
5788 7. Added expansion conditions match_domain, match_address, and match_local_
5789 part (NOT match_host).
5791 8. The placeholder for the Received: header didn't have a length field set.
5793 9. Added code to Exim itself and to exim_lock to test for a specific race
5794 condition that could lead to file corruption when using MBX delivery. The
5795 issue is with the lockfile that is created in /tmp. If this file is removed
5796 after a process has opened it but before that process has acquired a lock,
5797 there is the potential for a second process to recreate the file and also
5798 acquire a lock. This could lead to two Exim processes writing to the file
5799 at the same time. The added code performs the same test as UW imapd; it
5800 checks after acquiring the lock that its file descriptor still refers to
5801 the same named file.
5803 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5804 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5807 11. Added the smtp_active_hostname option. If used, this will typically be made
5808 to depend on the incoming interface address. Because $interface_address is
5809 not set up until the daemon has forked a reception process, error responses
5810 that can happen earlier (such as "too many connections") no longer contain
5813 12. If an expansion in a condition on a "warn" statement fails because a lookup
5814 defers, the "warn" statement is abandoned, and the next ACL statement is
5815 processed. Previously this caused the whole ACL to be aborted.
5817 13. Added the iplsearch lookup type.
5819 14. Added ident_timeout as a log selector.
5821 15. Added tls_certificate_verified as a log selector.
5823 16. Added a global option tls_require_ciphers (compare the smtp transport
5824 option of the same name). This controls incoming TLS connections.
5826 17. I finally figured out how to make tls_require_ciphers do a similar thing
5827 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5828 before starting the TLS session.
5830 18. Tabs are now shown as \t in -bP output.
5832 19. If the log selector return_path_on_delivery was set, Exim crashed when
5833 bouncing a message because it had too many Received: header lines.
5835 20. If two routers both had headers_remove settings, and the first one included
5836 a superfluous trailing colon, the final name in the first list and the
5837 first name in the second list were incorrectly joined into one item (with a
5838 colon in the middle).
5844 1. Added -C and -D options to the exinext utility, mainly to make it easier
5845 to include in the automated testing, but these could be helpful when
5846 multiple configurations are in use.
5848 2. The exinext utility was not formatting the output nicely when there was
5849 an alternate port involved in the retry record key, nor when there was a
5850 message id as well (for retries that were specific to a specific message
5851 and a specific host). It was also confused by IPv6 addresses, because of
5852 the additional colons they contain. I have fixed the IPv4 problem, and
5853 patched it up to do a reasonable job for IPv6.
5855 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5856 delivery, the log line now contains "pipelined" if PIPELINING was used.
5858 4. An SMTP transport process used to panic and die if the bind() call to set
5859 an explicit outgoing interface failed. This has been changed; it is now
5860 treated in the same way as a connect() failure.
5862 5. A reference to $sender_host_name in the part of a conditional expansion
5863 that was being skipped was still causing a DNS lookup. This no longer
5866 6. The def: expansion condition was not recognizing references to header lines
5867 that used bh_ and bheader_.
5869 7. Added the _cache feature to named lists.
5871 8. The code for checking quota_filecount in the appendfile transport was
5872 allowing one more file than it should have been.
5874 9. For compatibility with Sendmail, the command line option
5882 and sets the incoming protocol and host name (for trusted callers). The
5883 host name and its colon can be omitted when only the protocol is to be set.
5884 Note the Exim already has two private options, -pd and -ps, that refer to
5885 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5886 "s", but I don't think that's a major issue.
5888 10. A number of refactoring changes to the code, none of which should affect
5891 (a) The number of logging options was getting close to filling up the
5892 32-bit word that was used as a bit map. I have split them into two classes:
5893 those that are passed in the argument to log_write(), and those that are
5894 only ever tested independently outside of that function. These are now in
5895 separate 32-bit words, so there is plenty of room for expansion again.
5896 There is no change in the user interface or the logging behaviour.
5898 (b) When building, for example, log lines, the code previously used a
5899 macro that called string_cat() twice, in order to add two strings. This is
5900 not really sufficiently general. Furthermore, there was one instance where
5901 it was actually wrong because one of the argument was used twice, and in
5902 one call a function was used. (As it happened, calling the function twice
5903 did not affect the overall behaviour.) The macro has been replaced by a
5904 function that can join an arbitrary number of extra strings onto a growing
5907 (c) The code for expansion conditions now uses a table and a binary chop
5908 instead of a serial search (which was left over from when there were very
5909 few conditions). Also, it now recognizes conditions like "pam" even when
5910 the relevant support is not compiled in: a suitably worded error message is
5911 given if an attempt is made to use such a condition.
5913 11. Added ${time_interval:xxxxx}.
5915 12. A bug was causing one of the ddress fields not to be passed back correctly
5916 from remote delivery subprocesses. The field in question was not being
5917 subsequently used, so this caused to problems in practice.
5919 13. Added new log selectors queue_time and deliver_time.
5921 14. Might have fixed a bug in maildirsizefile handling that threw up
5922 "unexpected character" debug warnings, and recalculated the data
5923 unnecessarily. In any case, I expanded the warning message to give more
5926 15. Added the message "Restricted characters in address" to the statements in
5927 the default ACL that block characters like @ and % in local parts.
5929 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5930 Three changes have been made:
5932 (a) There was a serious bug; a negative response to MAIL caused the whole
5933 recipient domain to be cached as invalid, thereby blocking all messages
5934 to all local parts at the same domain, from all senders. This bug has
5935 been fixed. The domain is no longer cached after a negative response to
5936 MAIL if the sender used is not empty.
5938 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5941 (c) A new callout option, "use_sender" has been added for people who want
5942 the modified behaviour.
5948 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5951 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5952 indeed breaks things for older releases.
5954 3. Added additional logging to the case where there is a problem reading data
5955 from a filter that is running in a subprocess using a pipe, in order to
5956 try to track down a specific problem.
5958 4. Testing facility fudge: when running in the test harness and attempting
5959 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5960 getting "No route to host". Convert this to a timeout.
5962 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5965 6. Some OS don't have socklen_t but use size_t instead. This affects the
5966 fifth argument of getsockopt() amongst other things. This is now
5967 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5968 can be set for individual OS. I have set it for SunOS5, OSF1, and
5969 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5970 some earlier ones do not.
5972 7. Change 4.30/15 was not doing the test caselessly.
5974 8. The standard form for an IPv6 address literal was being rejected by address
5975 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5976 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5977 this, as well as the form without the "IPv6" on the front (but only when
5978 address literals are enabled, of course).
5980 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5982 10. Exim crashed if a message with an empty sender address specified by -f
5983 encountered a router with an errors_to setting. This could be provoked only
5984 by a command such as
5988 where an empty string was supplied; "<>" did not hit this bug.
5990 11. Installed PCRE release 4.5.
5992 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5993 remained set. It is now erased.
5995 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5996 times from message ids (which are base 36 rather than the normal 62).
5998 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5999 were being counted as actual protocol errors, and logged if the log
6000 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6001 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6002 and DATA following a set of rejected RCPTs do not count as protocol errors.
6003 In other words, Exim assumes they were pipelined, though this may not
6004 actually be the case. Of course, in all cases the client gets an
6005 appropriate error code.
6007 15. If a lookup fails in an ACL condition, a message about the failure may
6008 be available; it is used if testing the ACL cannot continue, because most
6009 such messages specify what the cause of the deferral is. However, some
6010 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6011 that caused an old message to be retained and used if a later statement
6012 caused a defer, replacing the real cause of the deferral.
6014 16. If an IP address had so many PTR records that the DNS lookup buffer
6015 was not large enough to hold them, Exim could crash while trying to process
6016 the truncated data. It now detects and logs this case.
6018 17. Further to 4.21/58, another change has been made: if (and only if) the
6019 first line of a message (the first header line) ends with CRLF, a bare LF
6020 in a subsequent header line has a space inserted after it, so as not to
6021 terminate the header.
6023 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6024 unnecessarily, used atoi() instead of strtol(), and didn't check the
6025 termination when getting file sizes from file names by regex.
6027 19. Completely re-implemented the support for maildirsize files, in the light
6028 of a number of problems with the previous contributed implementation
6029 (4.30/29). In particular:
6031 . If the quota is zero, the maildirsize file is maintained, but no quota is
6034 . If the maildir directory does not exist, it is created before any attempt
6035 to write a maildirsize file.
6037 . The quota value in the file is just a cache; if the quota is changed in
6038 the transport, the new value overrides.
6040 . A regular expression is available for excluding directories from the
6043 20. The autoreply transport checks the characters in options that define the
6044 message's headers; it allows continued headers, but it was checking with
6045 isspace() after an embedded newline instead of explicitly looking for a
6048 21. If all the "regular" hosts to which an address was routed had passed their
6049 expiry times, and had not reached their retry times, the address was
6050 bounced, even if fallback hosts were defined. Now Exim should go on to try
6053 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6054 equivalent code in the SMTP transport. Some hosts send humungous responses
6055 to HELO/EHLO, more than 1024 it seems.
6057 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6058 gives compiler warnings in some environments. I've now done it "properly",
6061 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6062 (because of problems with the built-in one) was declared to return uschar *
6063 instead of char *, causing compiler failure.
6065 25. Fixed a file descriptor leak when processing alias/forward files.
6067 26. Fixed a minor format string issue in dbfn.c.
6069 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6071 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6072 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6075 29. When a sender address is verified, it is cached, to save repeating the test
6076 when there is more than one recipient in a message. However, when the
6077 verification involves a callout, it is possible for different callout
6078 options to be set for different recipients. It is too complicated to keep
6079 track of this in the cache, so now Exim always runs a verification when a
6080 callout is required, relying on the callout cache for the optimization.
6081 The overhead is duplication of the address routing, but this should not be
6084 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6085 to be verified with callout=postmaster, and the main callout worked but the
6086 postmaster check failed, the verification correctly failed. However, if a
6087 subsequent RCPT command asked for sender verification *without* the
6088 postmaster check, incorrect caching caused this verification also to fail,
6091 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6092 it was not caching the DNS options (qualify_single, search_parents) that
6093 were used when the lookup failed. A subsequent lookup with different
6094 options therefore always gave the same answer, though there were cases
6095 where it should not have. (Example: a "domains = !$mx_any" option on a
6096 dnslookup router: the "domains" option is always processed without any
6097 widening, but the router might have qualify_single set.) Now Exim uses the
6098 cached value only when the same options are set.
6100 32. Added John Jetmore's "exipick" utility to the distribution.
6102 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6103 than a timeout (e.g. a certificate is required, and is not provided), an
6104 Exim server now closes the connection immediately. Previously it waited for
6105 the client to close - but if the client is SSL, it seems that they each
6106 wait for each other, leading to a delay before one of them times out.
6108 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6109 maintained 0.8.x compatibility because I don't think many are using it, and
6110 it is clearly obsolete.
6112 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6115 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6116 client certificate was expired. A simple patch fixes this, though I don't
6117 understand the full logic of why the verify callback is called multiple
6120 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6121 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6122 which causes problems with some clients (such as the Certicom SSL Plus
6123 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6124 disables the coutermeasure allowing Eudora to connect."
6126 38. Exim was not checking that a write() to a log file succeeded. This could
6127 lead to Bad Things if a log got too big, in particular if it hit a file
6128 size limit. Exim now panics and dies if it cannot write to a log file, just
6129 as it does if it cannot open a log file.
6131 39. Modified OS/Makefile-Linux so that it now contains
6133 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6135 The two -D definitions ensure that Exim is compiled with large file
6136 support, which makes it possible to handle log files that are bigger than
6139 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6140 instance) a domain was checked against a named list that involved a lookup,
6141 causing $domain_data to be set, then another domain was checked against the
6142 same list, then the first domain was re-checked, the value of $domain_data
6143 after the final check could be wrong. In particular, if the second check
6144 failed, it could be set empty. This bug probably also applied to
6147 41. The strip_trailing_dot option was not being applied to the address given
6148 with the -f command-line option.
6150 42. The code for reading a message's header from the spool was incrementing
6151 $received_count, but never initializing it. This meant that the value was
6152 incorrect (doubled) while delivering a message in the same process in which
6153 it was received. In the most common configuration of Exim, this never
6154 happens - a fresh exec is done - but it can happen when
6155 deliver_drop_privilege is set.
6157 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6158 it now includes up to 150 characters of the unexpected data in the log
6161 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6162 and building data strings. The size of both of these buffers was 10 000
6163 bytes - far larger than anybody would *ever* want, thought I. Needless to
6164 say, somebody hit the limit. I have increased the maximum line length to
6165 20 000 and the maximum data length of concatenated lines to 100 000. I have
6166 also fixed two bugs, because there was no checking on these buffers. Tsk,
6167 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6168 buffer is too small.
6170 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6171 lsearch lookups. Now it does.
6173 46. When parsing a route_list item in a manualroute router, a fixed-length
6174 buffer was used for the list of hosts. I made this 1024 bytes long,
6175 thinking that nobody would ever have a list of hosts that long. Wrong.
6176 Somebody had a whole pile of complicated expansion conditions, and the
6177 string was silently truncated, leading to an expansion error. It turns out
6178 that it is easier to change to an unlimited length (owing to other changes
6179 that have happened since this code was originally written) than to build
6180 structure for giving a limitation error. The length of the item that
6181 expands into the list of hosts is now unlimited.
6183 47. The lsearch lookup could not handle data where the length of text line was
6184 more than 4095 characters. Such lines were truncated, leading to shortened
6185 data being returned. It should now handle lines of any length.
6187 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6188 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6191 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6192 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6193 no longer happens in comment lines. A list of replacements is now placed
6194 at the head of all of the source files, except those whose only change is
6195 to replace PERL_COMMAND in the very first #! line.
6197 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6198 messages on the queue, with a bottom-up merge sort, using code contributed
6199 by Michael Haardt. This should make operations like -bp somewhat faster on
6200 large queues. It won't affect queue runners, except when queue_run_in_order
6203 51. Installed eximstats 1.31 in the distribution.
6205 52. Added support for SRV lookups to the dnslookup router.
6207 53. If an ACL referred to $message_body or $message_body_end, the value was not
6208 reset for any messages that followed in the same SMTP session.
6210 54. The store-handling optimization for building very long strings was not
6211 differentiating between the different store pools. I don't think this
6212 actually made any difference in practice, but I've tidied it.
6214 55. While running the routers to verify a sender address, $sender_address
6215 was still set to the sender address. This is wrong, because when routing to
6216 send a bounce to the sender, it would be empty. Therefore, I have changed
6217 it so that, while verifying a sender address, $sender_address is set to <>.
6218 (There is no change to what happens when verifying a recipient address.)
6220 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6221 target A or AAAA records (if not already returned) without resetting the
6222 qualify_single or search_parents options of the DNS resolver. These are
6223 inappropriate in this case because the targets of MX and SRV records must
6224 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6225 target that, when qualified, matched something in the local domain. These
6226 two options are now turned off when doing these lookups.
6228 57. It seems that at least some releases of Reiserfs (which does not have the
6229 concept of a fixed number of inodes) returns zero and not -1 for the
6230 number of available inodes. This interacted badly with check_spool_inodes,
6231 which assumed that -1 was the "no such thing" setting. What I have done is
6232 to check that the total number of inodes is greater than zero before doing
6233 the test of how many are available.
6235 58. When a "warn" ACL statement has a log_message modifier, the message is
6236 remembered, and not repeated. This is to avoid a lot of repetition when a
6237 message has many recipients that cause the same warning to be written.
6238 However, Exim was preserving the list of already written lines for an
6239 entire SMTP session, which doesn't seem right. The memory is now reset if a
6240 new message is started.
6242 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6243 debugging output unless log_rewrite was also set.
6245 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6246 of (int)(handle) when we know that handle contains (void *)(-1).
6248 61. The Exim daemon panic-logs an error return when it closes the incoming
6249 connection. However "connection reset by peer" seems to be common, and
6250 isn't really an error worthy of noting specially, so that particular error
6253 62. When Exim is trying to find all the local interfaces, it used to panic and
6254 die if the ioctl to get the interface flags failed. However, it seems that
6255 on at least one OS (Solaris 9) it is possible to have an interface that is
6256 included in the list of interfaces, but for which you get a failure error
6257 for this call. This happens when the interface is not "plumbed" into a
6258 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6259 failure of the "get flags" call assumes that the interface is down.
6261 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6262 makes life easier for people who are doing arithmetic on fields extracted
6263 from dates, where you often get leading zeros that should not be
6264 interpreted as octal.
6266 64. Added qualify_domain to the redirect router, to override the global
6269 65. If a pathologically long header line contained very many addresses (the
6270 report of this problem mentioned 10 000) and each of them was rewritten,
6271 Exim could use up a very large amount of memory. (It kept on making new
6272 copies of the header line as it rewrote, and never released the old ones.)
6273 At the expense of a bit more processing, the header rewriting function has
6274 been changed so that it no longer eats memory in this way.
6276 66. The generation of the Received: header has been moved from the time that a
6277 message starts to be received, to the time that it finishes. The timestamp
6278 in the Received: header should now be very close to that of the <= log
6279 line. There are two side-effects of this change:
6281 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6282 logged header lines no longer include the local Received: line, because
6283 it has not yet been created. The same applies to a copy of the message
6284 that is returned to a non-SMTP sender when a message is rejected.
6286 (b) When a filter file is tested using -bf, no additional Received: header
6287 is added to the test message. After some thought, I decided that this
6290 This change does not affect the value of $received_for. It is still set
6291 after address rewriting, but before local_scan() is called.
6293 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6295 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6296 gave an unhelpful panic error message, and a defer error. I have managed to
6297 change this behaviour so that it now rejects any supplied certificate,
6298 which seems right, as the list of acceptable certificates is empty.
6300 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6301 gave an unhelpful defer error. I have not managed to make this reject any
6302 supplied certificates, but the error message it gives is "no certificate
6303 supplied", which is not helpful.
6305 70. exigrep's output now also includes lines that are not associated with any
6306 message, but which match the given pattern. Implemented by a patch from
6307 Martin Sluka, which also tidied up the Perl a bit.
6309 71. Recipient callout verification, like sender verification, was using <> in
6310 the MAIL FROM command. This isn't really the right thing, since the actual
6311 sender may affect whether the remote host accepts the recipient or not. I
6312 have changed it to use the actual sender in the callout; this means that
6313 the cache record is now keyed on a recipient/sender pair, not just the
6314 recipient address. There doesn't seem to be a real danger of callout loops,
6315 since a callout by the remote host to check the sender would use <>.
6316 [SEE ABOVE: changed after hitting problems.]
6318 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6319 temporary errors. However, in the case of such a code being given after
6320 the end of a data transmission (i.e. after ".") Exim was failing to write
6321 a retry record for the message. (Yes, there was some broken host that was
6322 actually sending 8xx at this point.)
6324 73. An unknown lookup type in a host list could cause Exim to panic-die when
6325 the list was checked. (An example that provoked this was putting <; in the
6326 middle of a list instead of at the start.) If this happened during a DATA
6327 ACL check, a -D file could be left lying around. This kind of configuration
6328 error no longer causes Exim to die; instead it causes a defer error. The
6329 incident is still logged to the main and panic logs.
6331 74. Buglet left over from Exim 3 conversion. The message "too many messages
6332 in one connection" was written to the rejectlog but not the mainlog, except
6333 when address rewriting (yes!) was being logged.
6335 75. Added write_rejectlog option.
6337 76. When a system filter was run not as root (that is, when system_filter_user
6338 was set), the values of the $n variables were not being returned to the
6339 main process; thus, they were not subsequently available in the $sn
6342 77. Added +return_path_on_delivery log selector.
6344 78. A connection timeout was being treated differently from recipients deferred
6345 when testing hosts_max_try with a message that was older than the host's
6346 retry timeout. (The host should not be counted, thus allowing all hosts to
6347 be tried at least once before bouncing.) This may have been the cause of an
6348 occasionally reported bug whereby a message would remain on the queue
6349 longer than the retry timeout, but would be bounced if a delivery was
6350 forced. I say "may" because I never totally pinned down the problem;
6351 setting up timeout/retry tests is difficult. See also the next item.
6353 79. The ultimate address timeout was not being applied to errors that involved
6354 a combination of host plus message (for example, a timeout on a MAIL
6355 command). When an address resolved to a number of possible hosts, and they
6356 were not all tried for each delivery (e.g. because of hosts_max_try), a
6357 message could remain on the queue longer than the retry timeout.
6359 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6362 81. Fixed an obscure SMTP outgoing bug which required at least the following
6363 conditions: (a) there was another message waiting for the same server;
6364 (b) the server returned 5xx to all RCPT commands in the first message so
6365 that the message was not completed; (c) the server dropped the connection
6366 or gave a negative response to the RSET that Exim sends to abort the
6367 transaction. The observed case was a dropped connection after DATA that had
6368 been sent in pipelining mode. That is, the server had advertised PIPELINING
6369 but was not implementing it correctly. The effect of the bug was incorrect
6370 behaviour, such as trying another host, and this could lead to a crash.
6376 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6377 and daemon.c were passed as pointers to ints; they should have been
6378 pointers to socklen_t variables (which are typically unsigned ints).
6380 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6383 3. Fixed a really odd bug that affected only the testing scheme; patching a
6384 certain fixed string in the binary changed the value of another string that
6385 happened to be identical to the end of the original first string.
6387 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6388 name", it returns that address as the IP address. On some operating
6389 systems (e.g. Solaris), it also passes back the IP address string as the
6390 "host name". However, on others (e.g. Linux), it passes back an empty
6391 string. Exim wasn't checking for this, and was changing the host name to an
6392 empty string, assuming it had been canonicalized.
6394 5. Although rare, it is permitted to have more than one PTR record for a given
6395 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6396 all the names associated with an address, because they do in Solaris.
6397 However, it seems that they do not in Linux for data that comes from the
6398 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6399 I found this out when I moved to a new Linux workstation and tried to run
6400 the Exim test suite.
6402 To get round this problem I have changed the code so that it now does its
6403 own call to the DNS to look up PTR records when searching for a host name.
6404 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6405 addresses that are only in /etc/hosts are still found.
6407 This behaviour is, however, controlled by an option called host_lookup_
6408 order, which defaults to "bydns:byaddr". If people want to use the other
6409 order, or indeed, just use one or the other means of lookup, they can
6410 specify it in this variable.
6412 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6413 some operating systems, this comes back from gethostbyaddr() as an empty
6414 string, and this is what Exim used to test for. However, it seems that in
6415 other systems, "." is yielded. Exim now tests for this case too.
6417 7. The values of check_spool_space and check_log_space are now held internally
6418 as a number of kilobytes instead of an absolute number of bytes. If a
6419 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6420 kilobyte. This means that much larger values can be stored.
6422 8. Exim monitor: an attempt to get the action menu when not actually pointing
6423 at a message produces an empty menu entitled "No message selected". This
6424 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6425 no entries in it ("Shell widget menu has zero width and/or height"). So I
6426 have added a single, blank menu entry in this case.
6428 9. Added ${quote_local_part.
6430 10. MIME decoding is now applied to the contents of Subject: header lines when
6433 11. Now that a reference to $sender_host_address automatically causes a reverse
6434 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6435 host lookup before query-style lookups in lists that might use this
6436 variable. This has therefore been abolished, and the "net-" prefix is no
6437 longer necessary for query-style lookups.
6439 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6440 have been a typo for LFLAGS, so it has been changed.
6442 13. The install script calls Exim with "-C /dev/null" in order to find the
6443 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6444 to be output. However, since Exim outputs its version number before the
6445 error, it didn't break the script. It just looked ugly. I fixed this by
6446 always allowing "-C /dev/null" if the caller is root.
6448 14. Ignore overlarge ACL variable number when reading spool file - insurance
6449 against a later release with more variables having written the file.
6451 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6452 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6453 this, as well as the form without the "IPv6" on the front.
6455 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6456 OS/Makefile-Darwin file.
6458 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6459 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6461 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6462 variable. However, some OS use ints for inodes. I've added cast to long int
6463 to get rid of the compiler warning.
6465 19. I had forgotten to lock out "/../" in configuration file names when
6466 ALT_CONFIG_PREFIX was set.
6468 20. Routers used for verification do not need to specify transports. However,
6469 if such a router generated a host list, and callout was configured, Exim
6470 crashed, because it could not find a port number from the (non-existent)
6471 transport. It now assumes port 25 in this circumstance.
6473 21. Added the -t option to exigrep.
6475 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6476 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6477 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6478 them is set, LOOKUP_LSEARCH is forced.
6480 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6482 24. Added sender and host information to the "rejected by local_scan()" log
6483 line; previously there was no indication of these.
6485 25. Added .include_if_exists.
6487 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6488 Linux. It turns out that not all file systems support this. Apparently some
6489 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6490 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6491 not supported on the file descriptor, is now ignored when Exim is trying to
6492 sync a directory. This applies only to Linux.
6494 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6496 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6497 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6500 29. Added support for maildirsize files from supplied patch (modified a bit).
6502 30. The use of :fail: followed by an empty string could lead Exim to respond to
6503 sender verification failures with (e.g.):
6505 550 Verification failed for <xxx>
6506 550 Sender verify failed
6508 where the first response line was missing the '-' that indicates it is not
6509 the final line of the response.
6511 31. The loop for finding the name of the user that called Exim had a hardwired
6512 limit of 10; it now uses the value of finduser_retries, which is used for
6513 all other user lookups.
6515 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6518 33. Exim was neglecting to zero errno before one call of strtol() when
6519 expanding a string and expecting an integer value. On some systems this
6520 resulted in spurious "integer overflow" errors. Also, it was casting the
6521 result into an int without checking.
6523 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6524 did not work. The code looks as if it has *never* worked, though it appears
6525 to have been documented since at least release 1.62. I have made it work.
6527 35. The "timeout_DNS" error in retry rules, also documented since at least
6528 1.62, also never worked. As it isn't clear exactly what this means, and
6529 clearly it isn't a major issue, I have abolished the feature by treating it
6530 as "timeout", and writing a warning to the main and panic logs.
6532 36. The display of retry rules for -brt wasn't always showing the error code
6535 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6536 timeout_connect_A, timeout_connect_MX.
6538 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6539 to the empty sender.
6541 39. The daemon was not analysing the content of -oX till after it had closed
6542 stderr and disconnected from the controlling terminal. This meant that any
6543 syntax errors were only noted on the panic log, and the return code from
6544 the command was 0. By re-arranging the code a little, I've made the
6545 decoding happen first, so such errors now appear on stderr, and the return
6546 code is 1. However, the actual setting up of the sockets still happens in
6547 the disconnected process, so errors there are still only recorded on the
6550 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6551 connections (as happens on some IP stacks) was logged at start up time as
6552 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6553 it from "IPv6 and IPv4", which means that two separate sockets are being
6556 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6557 says whether AF_INET or AF_INET6 was passed as an argument.
6559 42. Exiwhat output was messed up when time zones were included in log
6562 43. Exiwhat now gives more information about the daemon's listening ports,
6563 and whether -tls-on-connect was used.
6565 44. The "port" option of the smtp transport is now expanded.
6567 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6568 silently ignored. Now an error message is written to the main and panic
6571 46. There's a new ACL modifier called "logwrite" which writes to a log file
6572 as soon as it is encountered.
6574 47. Added $local_user_uid and $local_user_gid at routing time.
6576 48. Exim crashed when trying to verify a sender address that was being
6579 49. Exim was recognizing only a space character after ".include". It now also
6580 recognizes a tab character.
6582 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6583 extracting the relevant information from the specification. The man page no
6584 longer contains scrambled data for the -d option, and I've added a section
6585 at the front about calling Exim under different names.
6587 51. Added "extra_headers" argument to the "mail" command in filter files.
6589 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6592 53. Installed eximstats 1.29.
6594 54. Added transport_filter_timeout as a generic transport option.
6596 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6597 Cc: header lines. This was required by RFC 822, but it not required by RFC
6600 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6601 incoming messages that did not have them. Now it does so only if the
6602 message originates locally, that is, if there is no associated remote host
6603 address. When Resent- header lines are present, this applies to the Resent-
6604 lines rather than the non-Resent- lines.
6606 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6607 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6609 58. Messages for configuration errors now include the name of the main
6610 configuration file - useful now that there may be more than one file in a
6611 list (.included file names were always shown).
6613 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6614 for those rare installations that do not start the daemon as root or run it
6615 setuid root. I've cut out the call to initgroups() if the daemon is not
6618 60. The Exim user and group can now be bound into the binary as text strings
6619 that are looked up at the start of Exim's processing.
6621 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6623 62. Added $mailstore_basename variable.
6625 63. Installed patch to sieve.c from Michael Haardt.
6627 64. When Exim failed to open the panic log after failing to open the main log,
6628 the original message it was trying to log was written to stderr and debug
6629 output, but if they were not available (the usual case in production), it
6630 was lost. Now it is written to syslog before the two lines that record the
6631 failures to open the logs.
6633 65. Users' Exim filters run in subprocesses under the user's uid. It is
6634 possible for a "deliver" command or an alias in a "personal" command to
6635 provoke an address rewrite. If logging of address rewriting is configured,
6636 this fails because the process is not running as root or exim. There may be
6637 a better way of dealing with this, but for the moment (because 4.30 needs
6638 to be released), I have disabled address rewrite logging when running a
6639 filter in a non-root, non-exim process.
6645 1. The buildconfig auxiliary program wasn't quoting the value set for
6646 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6647 not defined. This bug was masked in 4.22 by the effect that was fixed in
6650 2. Some messages that were rejected after a message id was allocated were
6651 shown as "incomplete" by exigrep. It no longer does this for messages that
6652 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6654 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6655 have allow_domain_literals set, the ID did not get logged in the <= line.
6656 Domain literals are now always recognized in Message-ID: header lines.
6658 4. The first argument for a ${extract expansion item is the key name or field
6659 number. Leading and trailing spaces in this item were not being ignored,
6660 causing some misleading effects.
6662 5. When deliver_drop_privilege was set, single queue runner processes started
6663 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6664 same command in the process it spins off) were not dropping privilege.
6666 6. When the daemon running as "exim" started a queue runner, it always
6667 re-executed Exim in the spun-off process. This is a waste of effort when
6668 deliver_drop_privilege is set. The new process now just calls the
6669 queue-runner function directly.
6675 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6678 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6679 ignored. Though the use of -oP was forcing the writing of a pid file, it
6680 was always written to the default place.
6682 3. If the message "no IP address found for host xxxx" is generated during
6683 incoming verification, it is now followed by identification of the incoming
6684 connection (so you can more easily find what provoked it).
6686 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6688 5. Added some features to "harden" Exim a bit more against certain attacks:
6690 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6691 be put in Local/Makefile. This is like the never_users runtime option,
6692 but it cannot be overridden. The default setting is "root".
6694 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6695 prefix string with which any file named in a -C command line option
6698 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6699 is retained for -C and -D only if the caller of Exim is root. Without
6700 it, the exim user may also use -C and -D and retain privilege.
6702 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6703 command line option is disabled.
6705 6. Macro names set by the -D option must start with an upper case letter, just
6706 like macro names defined in the configuration file.
6708 7. Added "dereference=" facility to LDAP.
6710 8. Two instances of the typo "uknown" in the source files are fixed.
6712 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6713 the Configure-Makefile script screwed up while processing it.
6715 10. Incorporated PCRE 4.4.
6717 11. The SMTP synchronization check was not operating right at the start of an
6718 SMTP session. For example, it could not catch a HELO sent before the client
6719 waited for the greeting. There is now a check for outstanding input at the
6720 point when the greeting is written. Because of the duplex, asynchronous
6721 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6722 way, but not yet received, when the check is performed.
6724 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6725 on TCP/IP sockets, because this apparently causes some broken clients to
6728 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6729 unchanged) from the Cygwin maintainer.
6731 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6732 was supported instead of testing for mbx. Effectively a typo.
6734 15. The spa authenticator server code was not checking that the input it
6735 received was valid base64.
6737 16. The debug output line for the "set" modifier in ACLs was not showing the
6738 name of the variable that was being set.
6740 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6742 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6743 lookup on its own. Something else was needed to trigger the lookup. For
6744 example, a match in host_lookup or the need for a host name in a host list.
6745 Now, if $sender_host_name is referenced and the host name has not yet been
6746 looked up, a lookup is performed. If the lookup fails, the variable remains
6747 empty, and $host_lookup_failed is set to "1".
6749 19. Added "eqi" as a case-independent comparison operator.
6751 20. The saslauthd authentication condition could segfault if neither service
6752 nor realm was specified.
6754 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6755 error message that was logged was misleading, and incoming SMTP
6756 connections were dropped. The message is now more accurate, and temporary
6757 errors are given to SMTP connections.
6759 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6760 (except RSET) with a 421 error, until QUIT is received. However, it was
6761 failing to send a response to QUIT.
6763 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6764 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6765 ACL is now run after the helo_try_verify_hosts code.
6767 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6768 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6769 case-independent, so other case variants are also recognized). Apparently
6770 some systems use these upper case variants.
6772 25. If more than two messages were waiting for the same host, and a transport
6773 filter was specified for the transport, Exim sent two messages over the
6774 same TCP/IP connection, and then failed with "socket operation on non-
6775 socket" when it tried to send the third.
6777 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6779 27. The extern definition of crypt16() in expand.c was not being excluded when
6780 the OS had its own crypt16() function.
6782 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6783 as a preferred synonym for return_size_limit, both as an option and as an
6786 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6788 30. Changed the default configuration ACL to relax the local part checking rule
6789 for addresses that are not in any local domains. For these addresses,
6790 slashes and pipe symbols are allowed within local parts, but the sequence
6791 /../ is explicitly forbidden.
6793 31. SPA server authentication was not clearing the challenge buffer before
6796 32. log_message in a "warn" ACL statement was writing to the reject log as
6797 well as to the main log, which contradicts the documentation and doesn't
6798 seem right (because no rejection is happening). So I have stopped it.
6800 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6801 However, I am unable to do any testing of this.
6803 34. Fixed an infelicity in the appendfile transport. When checking directories
6804 for a mailbox, to see if any needed to be created, it was accidentally
6805 using path names with one or more superfluous leading slashes; tracing
6806 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6808 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6809 contents are added to the log line that is written for every discarded
6810 recipient. (Previously a log_message setting was ignored.)
6812 36. The ${quote: operator now quotes the string if it is empty.
6814 37. The install script runs exim in order to find its version number. If for
6815 some reason other than non-existence or emptiness, which it checks, it
6816 could not run './exim', it was installing it with an empty version number,
6817 i.e. as "exim-". This error state is now caught, and the installation is
6820 38. An argument was missing from the function that creates an error message
6821 when Exim fails to connect to the socket for saslauthd authentication.
6822 This could cause Exim to crash, or give a corrupted message.
6824 39. Added isip, isip4, and isip6 to ${if conditions.
6826 40. The ACL variables $acl_xx are now saved with the message, and can be
6827 accessed later in routers, transports, and filters.
6829 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6830 strings in the file are not string-expanded.
6832 42. If a MAIL command specified a SIZE value that was too large to fit into an
6833 int variable, the check against message_size_limit failed. Such values are
6834 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6835 day this will have to be increased, but I don't think I want to be around
6836 when emails are that large.
6843 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6844 iconv() is not standard in FreeBSD.
6846 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6847 IPv6 enabled. The observed symptom was a segmentation fault on return from
6848 the function os_common_find_running_interfaces() in src/os.c.
6850 3. In the check_special_case() function in daemon.c I had used "errno" as an
6851 argument name, which causes warnings on some systems. This was basically a
6852 typo, since it was named "eno" in the comments!
6854 4. The code that waits for the clock to tick (at a resolution of some fraction
6855 of a second) so as to ensure message-id uniqueness was always waiting for
6856 at least one whole tick, when it could have waited for less. [This is
6857 almost certainly not relevant at current processor speeds, where it is
6858 unlikely to ever wait at all. But we try to future-proof.]
6860 5. The function that sleeps for a time interval that includes fractions of a
6861 second contained a race. It did not block SIGALRM between setting the
6862 timer, and suspending (a couple of lines later). If the interval was short
6863 and the sigsuspend() was delayed until after it had expired, the suspension
6864 never ended. On busy systems this could lead to processes getting stuck for
6867 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6868 process, before it forks any delivery processes. The open lookup caching
6869 mechanism meant that the open file or database connection was passed into
6870 the delivery process. The problem was that delivery processes always tidy
6871 up cached lookup data. This could cause a problem for the next delivery
6872 process started by the queue runner, because the external queue runner
6873 process does not know about the closure. So the next delivery process
6874 still has data in the lookup cache. In the case of a file lookup, there was
6875 no problem because closing a file descriptor in a subprocess doesn't affect
6876 the parent. However, if the lookup was caching a connection to a database,
6877 the connection was closed, and the second delivery process was likely to
6878 see errors such as "PGSQL: query failed: server closed the connection
6879 unexpectedly". The problem has been fixed by closing all cached lookups
6880 in a queue runner before running a delivery process.
6882 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6883 seem to have the "const" qualifier which it has on other OS. I've
6886 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6887 *of the same type* (client or server) with the same public name that an
6888 error should be diagnosed.
6890 9. When Exim looked up a host name for an IP address, but failed to find the
6891 original IP address when looking up the host name (a safety check), it
6892 output the message "<ip address> does not match any IP for NULL", which was
6893 confusing, to say the least. The bug was that the host name should have
6894 appeared instead of "NULL".
6896 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6897 user that is built into the binary, and the -C or -D options is used, root
6898 privilege is dropped before the configuration file is read. In addition,
6899 logging is switched to stderr instead of the normal log files. If the
6900 configuration then re-defines the Exim user, the unprivileged environment
6901 is probably not what is expected, so Exim logs a panic warning message (but
6904 However, if deliver_drop_privilege is set, the unprivileged state may well
6905 be exactly what is intended, so the warning has been cut out in that case,
6906 and Exim is allowed to try to write to its normal log files.
6912 1. smtp_return_error_details was not giving details for temporary sender
6913 or receiver verification errors.
6915 2. Diagnose a configuration error if two authenticators have the same public
6918 3. Exim used not to create the message log file for a message until the first
6919 delivery attempt. This could be confusing when incoming messages were held
6920 for policy or load reasons. The message log file is now created at the time
6921 the message is received, and an initial "Received" line is written to it.
6923 4. The automatically generated man page for command line options had a minor
6924 bug that caused no ill effects; however, a more serious problem was that
6925 the procedure for building the man page automatically didn't always
6926 operate. Consequently, release 4.20 contains an out-of-date version. This
6927 shouldn't happen again.
6929 5. When building Exim with embedded Perl support, the script that builds the
6930 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6931 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6933 6. The freeze_tell option was not being used for messages that were frozen on
6934 arrival, either by an ACL or by local_scan().
6936 7. Added the smtp_incomplete_transaction log selector.
6938 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6939 was accepting AUTH without a new EHLO.
6941 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6942 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6945 10. Logging of TCP/IP connections (when configured) now happens in the main
6946 daemon process instead of the child process, so that the TCP/IP connection
6947 count is more accurate (but it can never be perfect).
6949 11. The use of "drop" in a nested ACL was not being handled correctly in the
6950 outer ACL. Now, if condition failure induced by the nested "drop" causes
6951 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6952 or "require"), the connection is dropped.
6954 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6955 that yield "discard" can now be used with an "accept" or a "discard" verb,
6956 but an error is generated for any others (because I can't see a useful way
6957 to define what should happen).
6959 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6960 are now processed in the same way as lines in the Exim configuration file.
6961 In particular, continuation lines are supported.
6963 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6965 15. Added -ti meaning -t -i.
6967 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6968 domains, and warn by logging if others are found.
6970 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6971 returned from GIFCONF when Exim is trying to find the list of interfaces on
6972 a host. The code in os.c has been modified to copy each ifreq to an aligned
6973 structure in all cases.
6975 Also, in some cases, the returned ifreq's were being copied to a 'struct
6976 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6977 means the last couple of bytes of an IPv6 address could be chopped if the
6978 ifreq contained only a normal sockaddr (14 bytes storage).
6980 18. Named domain lists were not supported in the hosts_treat_as_local option.
6981 An entry such as +xxxx was not recognized, and was treated as a literal
6984 19. Ensure that header lines added by a DATA ACL are included in the reject log
6985 if the ACL subsequently rejects the message.
6987 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6988 MD5 (which is deprecated).
6990 21. When testing a filter file using -bf, Exim was writing a message when it
6991 took the sender from a "From " line in the message, but it was not doing so
6992 when it took $return_path from a Return-Path: header line. It now does.
6994 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6995 with a valid header line field name (a series of printing characters
6996 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6998 23. Changed "disc" in the source to "disk" to conform to the documentation and
6999 the book and for uniformity.
7001 24. Ignore Sendmail's -Ooption=value command line item.
7003 25. When execve() failed while trying to run a command in a pipe transport,
7004 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7005 could be confused with a return value of 69 from the command itself. This
7006 has been changed to 127, the value the shell returns if it is asked to run
7007 a non-existent command. The wording for the related log line suggests a
7008 non-existent command as the problem.
7010 26. If received_header_text expands to an empty string, do not add a Received:
7011 header line to the message. (Well, it adds a token one on the spool, but
7012 marks it "old" so that it doesn't get used or transmitted.)
7014 27. Installed eximstats 1.28 (addition of -nt option).
7016 28. There was no check for failure on the call to getsockname() in the daemon
7017 code. This can fail if there is a shortage of resources on the system, with
7018 ENOMEM, for example. A temporary error is now given on failure.
7020 29. Contrary to the C standard, it seems that in some environments, the
7021 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7022 program. Exim now does this explicitly; it affects the formatting of
7023 timestamps using strftime().
7025 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7026 complaints. I've now initialized all the variables, to avoid this.
7028 32. Header lines added by a system filter were not being "seen" during
7029 transport-time rewrites.
7031 33. The info_callback() function passed to OpenSSL is set up with type void
7032 (*)(SSL *, int, int), as described somewhere. However, when calling the
7033 function (actually a macro) that sets it up, the type void(*)() is
7034 expected. I've put in a cast to prevent warnings from picky compilers.
7036 34. If a DNS black list lookup found a CNAME record, but there were no A
7037 records associated with the domain it pointed at, Exim crashed.
7039 35. If a DNS black list lookup returned more than one A record, Exim ignored
7040 all but the first. It now scans all returned addresses if a particular IP
7041 value is being sought. In this situation, the contents of the
7042 $dnslist_value variable are a list of all the addresses, separated by a
7045 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7046 to accept a host name and all its aliases if the forward lookup for any of
7047 them yielded the IP address of the incoming connection. Now it accepts only
7048 those names whose forward lookup yields the correct IP address. Any other
7049 names are discarded. This closes a loophole whereby a rogue DNS
7050 administrator could create reverse DNS records to break through a
7051 wildcarded host restriction in an ACL.
7053 37. If a user filter or a system filter that ran in a subprocess used any of
7054 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7055 the wrong values were passed to the pipe command ($thisaddress had the
7056 value of $0, $0 had the value of $1, etc). This bug was introduced by
7057 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7059 38. Improved the line breaking for long SMTP error messages from ACLs.
7060 Previously, if there was no break point between 40 and 75 characters, Exim
7061 left the rest of the message alone. Two changes have been made: (a) I've
7062 reduced the minimum length to 35 characters; (b) if it can't find a break
7063 point between 35 and 75 characters, it looks ahead and uses the first one
7064 that it finds. This may give the occasional overlong line, but at least the
7065 remaining text gets split now.
7067 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7068 file descriptors might be low, and that setting 1000 would always raise it.
7069 It turns out that in some environments, the limit is already over 1000 and
7070 that lowering it causes trouble. So now Exim takes care not to decrease it.
7072 40. When delivering a message, the value of $return_path is set to $sender_
7073 address at the start of routing (routers may change the value). By an
7074 oversight, this default was not being set up when an address was tested by
7075 -bt or -bv, which affected the outcome if any router or filter referred to
7078 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7079 log only when "message" or "log_message" are set. However, if one of the
7080 conditions was an address verification, or a call to a nested ACL, the
7081 messages generated by the underlying test were being passed through. This
7082 no longer happens. The underlying message is available in $acl_verify_
7083 message for both "message" and "log_message" expansions, so it can be
7084 passed through if needed.
7086 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7087 new expansion $bh_ to give the encoded byte string without charset
7088 translation. Translation happens only if iconv() is available; HAVE_ICONV
7089 indicates this at build time. HEADERS_CHARSET gives the charset to
7090 translate to; headers_charset can change it in the configuration, and
7091 "headers charset" can change it in an individual filter file.
7093 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7094 that creates RFC 2047 encoded "words" labels them as that charset instead
7095 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7096 expansion operator; (ii) when Exim creates a From: line for a local
7097 message; (iii) when a header line is rewritten to include a "phrase" part.
7099 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7100 buggy, causing it to skip the first lines of messages whose message ID
7101 ended in 'D'. This would not have bitten before Exim release 4.14, because
7102 message IDs were unlikely to end in 'D' before then. The effect was to have
7103 incorrect size information for certain domains.
7105 45. #include "config.h" was missing at the start of the crypt16.c module. This
7106 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7109 46. If there was a timeout during a "random" callout check, Exim treated it as
7110 a failure of the random address, and carried on sending RSET and the real
7111 address. If the delay was just some slowness somewhere, the response to the
7112 original RCPT would be taken as a response to RSET and so on, causing
7113 mayhem of various kinds.
7115 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7116 when I implemented it. It didn't allow for the fact that some option values
7117 may legitimately be negative (e.g. size_addition), and it didn't even do
7118 the right test for positive values.
7120 48. Domain names in DNS records are case-independent. Exim always looks them up
7121 in lower case. Some resolvers return domain names in exactly the case they
7122 appear in the zone file, that is, they may contain uppercase letters. Not
7123 all resolvers do this - some return always lower case. Exim was treating a
7124 change of case by a resolver as a change of domain, similar to a widening
7125 of a domain abbreviation. This triggered its re-routing code and so it was
7126 trying to route what was effectively the same domain again. This normally
7127 caused routing to fail (because the router wouldn't handle the domain
7128 twice). Now Exim checks for this case specially, and just changes the
7129 casing of the domain that it ultimately uses when it transmits the message
7132 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7135 50. If a filter generated a file delivery with a non-absolute name (possible if
7136 no home directory exists for the router), the forbid_file option was not
7139 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7140 the existing equality matching.
7142 52. Exim was using ints instead of ino_t variables in some places where it was
7143 dealing with inode numbers.
7145 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7146 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7147 called TMPDIR, and if it finds it is different, it changes its value.
7149 54. The smtp_printf() function is now made available to local_scan() so
7150 additional output lines can be written before returning. There is also an
7151 smtp_fflush() function to enable the detection of a dropped connection.
7152 The variables smtp_input and smtp_batched_input are exported to
7155 55. Changed the default runtime configuration: the message "Unknown user"
7156 has been removed from the ACL, and instead placed on the localuser router,
7157 using the cannot_route_message feature. This means that any verification
7158 failures that generate their own messages won't get overridden. Similarly,
7159 the "Unrouteable address" message that was in the ACL for unverifiable
7160 relay addresses has also been removed.
7162 56. Added hosts_avoid_esmtp to the smtp transport.
7164 57. The exicyclog script was not checking for the esoteric option
7165 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7166 will work only if exicyclog is run under the appropriate euid.
7168 58. Following a discussion on the list, the rules by which Exim recognises line
7169 endings on incoming messages have been changed. The -dropcr and drop_cr
7170 options are now no-ops, retained only for backwards compatibility. The
7171 following line terminators are recognized: LF CRLF CR. However, special
7172 processing applies to CR:
7174 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7175 nor a local message in the state where . is a terminator.
7177 (ii) If a bare CR is encountered in a header line, an extra space is added
7178 after the line terminator so as not to end the header. The reasoning
7179 behind this is that bare CRs in header lines are most likely either
7180 to be mistakes, or people trying to play silly games.
7182 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7183 was being incorrectly given as 18 bytes larger than it should have been.
7184 This is a VOB (very old bug).
7186 60. This may never have affected anything current, but just in case it has:
7187 When the local host is found other than at the start of a list of hosts,
7188 the local host, those with the same MX, and any that follow, are discarded.
7189 When the list in question was part of a longer list of hosts, the following
7190 hosts (not currently being processed) were also being discarded. This no
7191 longer happens. I'm not sure if this situation could ever has previously
7194 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7197 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7198 header. This is something that is recommended in a new Internet Draft, and
7199 is something that is documented as being done by Sendmail. There are two
7200 possible values. For messages generated by the autoreply transport, Exim
7203 Auto-Submitted: auto-replied
7205 whereas for all other generated messages (e.g. bounces) it adds
7207 Auto-Submitted: auto-generated
7209 63. The "personal" condition in filters now includes a test for the
7210 Auto-Submitted: header. If it contains the string "auto-" the message it
7211 not considered personal.
7213 64. Added rcpt_include_affixes as a generic transport option.
7215 65. Added queue_only_override (default true).
7217 66. Added the syslog_duplication option.
7219 67. If what should have been the first header line of a message consisted of
7220 a space followed by a colon, Exim was mis-interpreting it as a header line.
7221 It isn't of course - it is syntactically invalid and should therefore be
7222 treated as the start of the message body. The misbehaviour could have
7223 caused a number of strange effects, including loss of data in subsequent
7224 header lines, and spool format errors.
7226 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7227 client host had authenticated. This control can now be exercised by an ACL
7228 for more flexibility.
7230 69. By default, callouts do not happen when testing with -bh. There is now a
7231 variant, -bhc, which does actually run the callout code, including
7232 consulting and updating the callout cache.
7234 70. Added support for saslauthd authentication, courtesy of Alexander
7237 71. If statvfs() failed on the spool or log directories while checking their
7238 size for availability, Exim confusingly gave the error "space shortage".
7239 Furthermore, in debugging mode it crashed with a floating point exception.
7240 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7241 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7242 serious problem, Exim now writes to the main and panic logs when this
7243 happens, with details of the failure. It then refuses to accept the
7244 incoming message, giving the message "spool directory problem" or "log
7245 directory problem" with a 421 code for SMTP messages.
7247 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7248 0, 1, and 2 exist, because some OS complain for execs without them (see
7249 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7250 descriptors. However, the code omitted to check that the open succeeded,
7251 causing mysterious errors if for some reason the permissions on /dev/null
7252 got screwed. Now Exim writes a message to the main and panic logs, and
7253 bombs out if it can't open /dev/null.
7255 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7256 interact so that it is all more flexible. It is supposed to remain
7257 backwards compatible. Also added extra_local_interfaces.
7259 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7260 to bomb out with an assertion failure - to the client this appears as a
7261 connection drop. This problem occurs in the part of the code that was taken
7262 from the Samba project. Fortunately, the assertion is in a very simple
7263 function, so I have fixed this by reproducing the function inline in the
7264 one place where it is called, and arranging for authentication to fail
7265 instead of killing the process with assert().
7267 75. The SPA client code was not working when the server requested OEM rather
7268 than Unicode encoding.
7270 76. Added code to make require_files with a specific uid setting more usable in
7271 the case where statting the file as root fails - usually a non-root-mounted
7272 NFS file system. When this happens and the failure is EACCES, Exim now
7273 forks a subprocess and does the per-uid checking as the relevant uid.
7275 77. Added process_log_path.
7277 78. If log_file_path was not explicitly set, a setting of check_log_space or
7278 check_log_inodes was ignored.
7280 79. If a space check for the spool or log partitions fails, the incident is now
7281 logged. Of course, in the latter case the data may get lost...
7283 80. Added the %p formatting code to string_format() so that it can be used to
7284 print addresses in debug_print(). Adjusted all the address printing in the
7285 debugging in store.c to use %p rather than %d.
7287 81. There was a concern that a line of code in smtp_in.c could overflow a
7288 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7289 initially expressed, the concern was not well-founded, because trailing
7290 spaces are removed early. However, if the trailing spaces were followed by
7291 a NULL, they did not get removed, so the overflow was possible. Two fixes
7294 (a) I re-wrote the offending code in a cleaner fashion.
7295 (b) If an incoming SMTP command contains a NULL character, it is rejected
7298 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7299 runs initgroups(), so that if the Exim user is in any additional groups,
7300 they will be used during message reception.
7306 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob