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.
137 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
138 msg:fail:internal event. Developement by Matthias Kurz.
140 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
141 far too small for todays use of crypto signatures stored there. Go all
142 the way to the max DNS message size of 64kB, even though this might be
143 overmuch for IOT constrained device use.
145 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
146 copy a string over itself. The library routine is documented as not
147 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
153 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
154 When redis_servers is set to a list of > 1 element, and the Redis servers
155 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
156 case of MOVED into a DEFER case instead, thus moving the query onto the
157 next server in the list. For a cluster of N elements, all N servers must
158 be defined in redis_servers.
160 GF/02 Catch and remove uninitialized value warning in exiqsumm
161 Check for existence of @ARGV before looking at $ARGV[0]
163 JH/01 Replace the store_release() internal interface with store_newblock(),
164 which internalises the check required to safely use the old one, plus
165 the allocate and data copy operations duplicated in both (!) of the
166 extant use locations.
168 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
169 modifier. This matches the restriction on the commandline.
171 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
172 Previously only the last row was returned.
174 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
175 we assumed that tags in the header were well-formed, and parsed the
176 element content after inspecting only the first char of the tag.
177 Assumptions at that stage could crash the receive process on malformed
180 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
181 While running the DKIM ACL we operate on the Permanent memory pool so that
182 variables created with "set" persist to the DATA ACL. Also (at any time)
183 DNS lookups that fail create cache records using the Permanent pool. But
184 expansions release any allocations made on the current pool - so a dnsdb
185 lookup expansion done in the DKIM ACL releases the memory used for the
186 DNS negative-cache, and bad things result. Solution is to switch to the
187 Main pool for expansions.
188 While we're in that code, add checks on the DNS cache during store_reset,
189 active in the testsuite.
190 Problem spotted, and debugging aided, by Wolfgang Breyha.
192 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
193 When none of the hosts presented to a transport match an already-open
194 connection, close it and proceed with the list. Previously we would
195 queue the message. Spotted by Lena with Yahoo, probably involving
198 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
199 Previously a spurious "250 OK id=" response was appended to the proper
202 JH/08 The "support for" informational output now, which built with Content
203 Scanning support, has a line for the malware scanner interfaces compiled
204 in. Interface can be individually included or not at build time.
206 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
207 by the template makefile "src/EDITME". The "STREAM" support for an older
208 ClamAV interface method is removed.
210 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
211 rows affected is given instead).
213 JH/11 The runtime Berkeley DB library version is now additionally output by
214 "exim -d -bV". Previously only the compile-time version was shown.
216 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
217 SMTP connection. Previously, when one had more receipients than the
218 first, an abortive onward connection was made. Move to full support for
219 multiple onward connections in sequence, handling cutthrough connection
220 for all multi-message initiating connections.
222 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
223 routers. Previously, a multi-recipient message would fail to match the
224 onward-connection opened for the first recipient, and cause its closure.
226 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
227 a timeout on read on a GnuTLS initiating connection, resulting in the
228 initiating connection being dropped. This mattered most when the callout
229 was marked defer_ok. Fix to keep the two timeout-detection methods
232 JH/15 Relax results from ACL control request to enable cutthrough, in
233 unsupported situations, from error to silently (except under debug)
234 ignoring. This covers use with PRDR, frozen messages, queue-only and
237 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
239 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
240 metadata, resulting in a crash in free().
242 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
243 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
244 Broken also in d185889f4, with init system revamp.
246 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
247 Previously we abruptly closed the connection after reading a malware-
248 found indication; now we go on to read the "scan ok" response line,
251 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
252 ACL. Previously, a crash would result.
254 JH/19 Speed up macro lookups during configuration file read, by skipping non-
255 macro text after a replacement (previously it was only once per line) and
256 by skipping builtin macros when searching for an uppercase lead character.
258 JH/20 DANE support moved from Experimental to mainline. The Makefile control
259 for the build is renamed.
261 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
262 was allocated for every new TLS startup, meaning one per message. Fix
263 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
265 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
266 reported the original. Fix to report (as far as possible) the ACL
267 result replacing the original.
269 JH/23 Fix memory leak during multi-message connections using STARTTLS under
270 OpenSSL. Certificate information is loaded for every new TLS startup,
271 and the resources needed to be freed.
273 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
275 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
278 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
279 DATA response info to the (existing) per-recipient response info for
280 the "C=" log element. It can have useful tracking info from the
281 destination system. Patch from Simon Arlott.
283 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
284 length value. Previously this would segfault.
286 HS/02 Support Avast multiline protoocol, this allows passing flags to
287 newer versions of the scanner.
289 JH/28 Ensure that variables possibly set during message acceptance are marked
290 dead before release of memory in the daemon loop. This stops complaints
291 about them when the debug_store option is enabled. Discovered specifically
292 for sender_rate_period, but applies to a whole set of variables.
293 Do the same for the queue-runner and queue-list loops, for variables set
294 from spool message files. Do the same for the SMTP per-message loop, for
295 certain variables indirectly set in ACL operations.
297 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
298 as a multi-recipient message from a mailinglist manager). The coding had
299 an arbitrary cutoff number of characters while checking for more input;
300 enforced by writing a NUL into the buffer. This corrupted long / fast
301 input. The problem was exposed more widely when more pipelineing of SMTP
302 responses was introduced, and one Exim system was feeding another.
303 The symptom is log complaints of SMTP syntax error (NUL chars) on the
304 receiving system, and refused recipients seen by the sending system
305 (propating to people being dropped from mailing lists).
306 Discovered and pinpointed by David Carter.
308 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
309 replaced by the ${authresults } expansion.
311 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
313 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
314 allows proper process termination in container environments.
316 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
317 Previously the "final dot" had a newline after it; ensure it is CR,LF.
319 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
320 and "err_perm", deprecated since 4.83 when the RFC-defined words
321 "temperror" and "permerror" were introduced.
323 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
324 transport-filters or DKIM-signing. The restriction was lost in the
325 consolidation of verify-callout and delivery SMTP handling.
326 Extend the restriction to also cover ARC-signing.
328 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
329 in defer=pass mode supply a 450 to the initiator. Previously the message
332 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
333 tls_require_ciphers is used as before.
335 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
336 "pass_unscanned". Only tmpfails from the scanner are written to
337 the paniclog, as they may require admin intervention (permission
338 denied, license issues). Other scanner errors (like decompression
339 bombs) do not cause a paniclog entry.
341 JH/36 Fix reinitialisation of DKIM logging variable between messages.
342 Previously it was possible to log spurious information in receive log
345 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
346 triggered odd behaviour from Outlook Express clients.
348 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
351 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
352 since the IETF WG has not yet settled on that versus the original
353 "bare" representation.
355 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
356 Previously the millisecond value corrupted the output.
357 Fix also for syslog_pid=no and log_selector +pid, for which the pid
358 corrupted the output.
364 JH/01 Rework error string handling in TLS interface so that the caller in
365 more cases is responsible for logging. This permits library-sourced
366 string to be attached to addresses during delivery, and collapses
367 pairs of long lines into single ones.
369 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
370 during configuration. Wildcards are allowed and expanded.
372 JH/02 Rework error string handling in DKIM to pass more info back to callers.
373 This permits better logging.
375 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
376 do not close it down and have the child transport start it up again on
377 the passed-on TCP connection. Instead, proxy the child (and any
378 subsequent ones) for TLS via a unix-domain socket channel. Logging is
379 affected: the continued delivery log lines do not have any DNSSEC, TLS
380 Certificate or OCSP information. TLS cipher information is still logged.
382 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
383 identical IP addresses on different listening ports. Will also affect
386 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
387 add noisy ifdef guards to special-case this sillyness.
388 Patch from Bernd Kuhls.
390 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
391 than 255 are no longer allowed.
393 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
394 Disable the session-cache too, which might reduce our load. Since we
395 currrectly use a new context for every connection, both as server and
396 client, there is no benefit for these.
397 GnuTLS appears to not support tickets server-side by default (we don't
398 call gnutls_session_ticket_enable_server()) but client side is enabled
399 by default on recent versions (3.1.3 +) unless the PFS priority string
402 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
403 <https://reproducible-builds.org/specs/source-date-epoch/>.
405 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
406 the check for any unsuccessful recipients did not notice the limit, and
407 erroneously found still-pending ones.
409 JH/08 Pipeline CHUNKING command and data together, on kernels that support
410 MSG_MORE. Only in-clear (not on TLS connections).
412 JH/09 Avoid using a temporary file during transport using dkim. Unless a
413 transport-filter is involved we can buffer the headers in memory for
414 creating the signature, and read the spool data file once for the
415 signature and again for transmission.
417 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
418 4.77 as the kernel support then wasn't solid, having issues in 64bit
419 mode. Now, it's been long enough. Add support for FreeBSD also.
421 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
422 case where the routing stage had gathered several addresses to send to
423 a host before calling the transport for the first, we previously failed
424 to close down TLS in the old transport process before passing the TCP
425 connection to the new process. The new one sent a STARTTLS command
426 which naturally failed, giving a failed delivery and bloating the retry
427 database. Investigation and fix prototype from Wolfgang Breyha.
429 JH/12 Fix check on SMTP command input synchronisation. Previously there were
430 false-negatives in the check that the sender had not preempted a response
431 or prompt from Exim (running as a server), due to that code's lack of
432 awareness of the SMTP input buffering.
434 PP/04 Add commandline_checks_require_admin option.
435 Exim drops privileges sanely, various checks such as -be aren't a
436 security problem, as long as you trust local users with access to their
437 own account. When invoked by services which pass untrusted data to
438 Exim, this might be an issue. Set this option in main configuration
439 AND make fixes to the calling application, such as using `--` to stop
442 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
443 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
444 responses to those, into a single TLS record each way (this usually means
445 a single packet). As a side issue, smtp_enforce_sync now works on TLS
448 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
449 affects you only if you're dancing at the edge of the param size limits.
450 If you are, and this message makes sense to you, then: raise the
451 configured limit or use OpenSSL 1.1. Nothing we can do for older
454 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
455 cmdline element to get the documented default one. Previously it was
458 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
459 are unsuable for non-message-specific reasons - eg. connection timeout,
460 banner-time rejection.
462 JH/16 Fix logging of delivery remote port, when specified by router, under
465 PP/06 Repair manualroute's ability to take options in any order, even if one
466 is the name of a transport.
469 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
471 JH/17 Change the list-building routines interface to use the expanding-string
472 triplet model, for better allocation and copying behaviour.
474 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
475 Previously it was constructed the first time a possibly-matching string
476 was met in the configuration file input during startup; now it is done
479 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
480 compatible one, to avoid the (poorly documented) possibility of a config
481 file in the working directory redirecting the DB files, possibly correpting
482 some existing file. CVE-2017-10140 assigned for BDB.
484 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
485 cache-hot. Previously, although the result was properly cached, the
486 initial verify call returned a defer.
488 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
489 the main verify for receipient in uncached-mode.
491 JH/22 Retire historical build files to an "unsupported" subdir. These are
492 defined as "ones for which we have no current evidence of testing".
494 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
495 if present. Previously it was ignored.
497 JH/24 Start using specified-initialisers in C structure init coding. This is
498 a C99 feature (it's 2017, so now considered safe).
500 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
501 if was a fixed-sized field and bitmask ops via macros; it is now more
504 PP/07 GitHub PR 56: Apply MariaDB build fix.
505 Patch provided by Jaroslav Škarvada.
507 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
508 during Coverity cleanups [4.87 JH/47]
509 Diagnosis and fix provided by Michael Fischer v. Mollard.
511 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
512 the right size to place the terminating semicolon on its own folded
513 line, the header hash was calculated to an incorrect value thanks to
514 the (relaxed) space the fold became.
516 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
517 and confused the parent.
519 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
520 which could crash as a result. This could lead to undeliverable messages.
522 JH/28 Logging: "next input sent too soon" now shows where input was truncated
525 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
526 matters on fast-turnover and PID-randomising systems, which were getting
527 out-of-order delivery.
529 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
530 a possibly-overlapping copy. The symptom was that "Remote host closed
531 connection in response to HELO" was logged instead of the actual 4xx
534 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
535 Previously only that bufferd was discarded, resulting in SYMTP command
538 JH/32 DKIM: when a message has multiple signatures matching an identity given
539 in dkim_verify_signers, run the dkim acl once for each. Previously only
540 one run was done. Bug 2189.
542 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
543 "panic the current process" to "deliberately defer". The panic log is
544 still written with the problem list name; the mail and reject logs now
545 get a temp-reject line for the message that was being handled, saying
546 something like "domains check lookup or other defer". The SMTP 451
547 message is still "Temporary local problem".
549 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
550 A crafted sequence of BDAT commands could result in in-use memory beeing
551 freed. CVE-2017-16943.
553 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
554 from SMTP input. Previously it was always done; now only done for DATA
555 and not BDAT commands. CVE-2017-16944.
557 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
558 to the message (such as an overlong header line). Previously this was
559 not done and we did not exit BDAT mode. Followon from the previous item
560 though a different problem.
566 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
567 than -2003 did; needs libidn2 in addition to libidn.
569 JH/02 The path option on a pipe transport is now expanded before use.
571 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
572 Patch provided by "Björn", documentation fix added too.
574 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
575 missing a wire-to-host endian conversion.
577 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
578 close after a BDAT command line could be taken as a following command,
579 giving a synch failure. Fix by only checking for synch immediately
580 before acknowledging the chunk.
582 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
583 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
584 macro. Patches provided by Josh Soref.
586 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
587 Previously we did not; the RFC seems ambiguous and VRFY is not listed
588 by IANA as a service extension. However, John Klensin suggests that we
591 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
592 the dkim code may be unix-mode line endings rather than smtp wire-format
593 CRLF, so prepend a CR to any bare LF.
595 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
596 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
598 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
599 the first were themselves being wrongly included in the feed into dkim
600 processing; with most chunk sizes in use this resulted in an incorrect
601 body hash calculated value.
603 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
604 DKIM signature block, for verification. Although advised against by
605 standards it is specifically not ruled illegal.
607 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
609 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
610 missing a body hash (the bh= tag).
612 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
613 It seems that HAProxy sends the Proxy Protocol information in clear and
614 only then does a TLS startup, so do the same.
616 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
617 TCP connections (such as for Spamd) unless the daemon successfully set
618 Fast Open mode on its listening sockets. This fixes breakage seen on
619 too-old kernels or those not configured for Fast Open, at the cost of
620 requiring both directions being enabled for TFO, and TFO never being used
621 by non-daemon-related Exim processes.
623 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
624 endings, at least on the first header line. Try to canonify any that get
625 past that check, despite the cost.
627 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
628 now limited to an arbitrary five deep, while parsing addresses with the
629 strip_excess_angle_brackets option enabled.
631 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
632 instead leave the unprompted TLS handshake in socket buffer for the
633 TLS library to consume.
635 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
637 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
639 JH/16 Drop variables when they go out of scope. Memory management drops a whole
640 region in one operation, for speed, and this leaves assigned pointers
641 dangling. Add checks run only under the testsuite which checks all
642 variables at a store-reset and panics on a dangling pointer; add code
643 explicitly nulling out all the variables discovered. Fixes one known
644 bug: a transport crash, where a dangling pointer for $sending_ip_address
645 originally assigned in a verify callout, is re-used.
647 PP/06 Drop '.' from @INC in various Perl scripts.
649 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
651 PP/08 Reduce a number of compilation warnings under clang; building with
652 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
653 should be warning-free.
655 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
657 HS/01 Fix portability problems introduced by PP/08 for platforms where
658 realloc(NULL) is not equivalent to malloc() [SunOS et al].
660 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
661 chunk. This allows us to accept broken chunked messages. We need a more
662 general solution here.
664 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
665 already-broken messages in the queue.
667 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
669 JH/19 Fix reference counting bug in routing-generated-address tracking.
675 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
676 supports it and a size is available (ie. the sending peer gave us one).
678 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
679 years of being deprecated). The replacements are the ACLs
680 acl_smtp_mime and acl_not_smtp_mime.
682 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
683 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
684 or even an in-clear connection were permitted. Now, if the host lookup
685 was dnssec and dane was requested then the host is only used if the
686 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
687 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
688 if one fails this test.
689 This means that a poorly-configured remote DNS will make it incommunicado;
690 but it protects against a DNS-interception attack on it.
692 JH/04 Bug 1810: make continued-use of an open smtp transport connection
693 non-noisy when a race steals the message being considered.
695 JH/05 If main configuration option tls_certificate is unset, generate a
696 self-signed certificate for inbound TLS connections.
698 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
699 in rewrites and routers.
701 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
702 and logged a warning sing 4.83; now they are a configuration file error.
704 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
705 (lacking @domain). Apply the same qualification processing as RCPT.
707 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
709 JH/10 Support ${sha256:} applied to a string (as well as the previous
712 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
713 a cutthrough deliver is pending, as we always want to make a connection.
714 This also avoids re-routing the message when later placing the cutthrough
715 connection after a verify cache hit.
716 Do not update it with the verify result either.
718 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
719 when routing results in more than one destination address.
721 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
722 signing (which inhibits the cutthrough capability). Previously only
723 the presence of an option was tested; now an expansion evaluating as
724 empty is permissible (obviously it should depend only on data available
725 when the cutthrough connection is made).
727 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
728 the relevant preceding SMTP command did not note the pipelining mode.
730 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
731 Previously they were not counted.
733 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
734 as one having no matching records. Previously we deferred the message
735 that needed the lookup.
737 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
738 distinguished as "(=".
740 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
741 for missing MX records. Previously it only worked for missing A records.
743 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
745 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
746 after the data-go-ahead and data-ack. Patch from Jason Betts.
748 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
749 even for a "none" policy. Patch from Tony Meyer.
751 JH/22 Fix continued use of a connection for further deliveries. If a port was
752 specified by a router, it must also match for the delivery to be
755 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
756 When one of the recipients of a message was unsuitable for the connection
757 (has no matching addresses), we lost track of needing to mark it
758 deferred. As a result mail would be lost.
760 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
762 JH/25 Decoding ACL controls is now done using a binary search; the source code
763 takes up less space and should be simpler to maintain. Merge the ACL
764 condition decode tables also, with similar effect.
766 JH/26 Fix problem with one_time used on a redirect router which returned the
767 parent address unchanged. A retry would see the parent address marked as
768 delivered, so not attempt the (identical) child. As a result mail would
771 JH/27 Fix a possible security hole, wherein a process operating with the Exim
772 UID can gain a root shell. Credit to http://www.halfdog.net/ for
773 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
776 JH/28 Enable {spool,log} filesystem space and inode checks as default.
777 Main config options check_{log,spool}_{inodes,space} are now
778 100 inodes, 10MB unless set otherwise in the configuration.
780 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
781 Previously it only applied to the main-section connection policy
784 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
786 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
787 by me. Added RFC7919 DH primes as an alternative.
789 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
790 are not in the system include path.
792 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
793 GnuTLS, when a session startup failed (eg because the client disconnected)
794 Exim did stdio operations after fclose. This was exposed by a recent
795 change which nulled out the file handle after the fclose.
797 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
798 signed directly by the cert-signing cert, rather than an intermediate
799 OCSP-signing cert. This is the model used by LetsEncrypt.
801 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
803 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
804 an incoming connection.
806 HS/02 Bug 1802: Do not half-close the connection after sending a request
809 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
810 fallback to "prime256v1".
812 JH/34 SECURITY: Use proper copy of DATA command in error message.
813 Could leak key material. Remotely exploitable. CVE-2016-9963.
819 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
820 and 3.4.4 - once the server is enabled to respond to an OCSP request
821 it does even when not requested, resulting in a stapling non-aware
822 client dropping the TLS connection.
824 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
825 support variable-length bit vectors. No functional change.
827 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
828 The I= interface field on outgoing lines is now after the H= remote
829 host field, same as incoming lines. There is a separate
830 outgoing_interface log selector which allows you to disable the
833 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
834 If not running log_selector +smtp_connection the mainlog would be held
835 open indefinitely after a "too many connections" event, including to a
836 deleted file after a log rotate. Leave the per net connection logging
837 leaving it open for efficiency as that will be quickly detected by the
838 check on the next write.
840 HS/01 Bug 1671: Fix post transport crash.
841 Processing the wait-<transport> messages could crash the delivery
842 process if the message IDs didn't exist for some reason. When
843 using 'split_spool_directory=yes' the construction of the spool
844 file name failed already, exposing the same netto behaviour.
846 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
847 mime_regex ACL conditions.
849 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
850 to DSN fail messages (bounces): remote IP, remote greeting, remote response
851 to HELO, local diagnostic string.
853 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
854 log line to debug. Even when configured with a tls authenticator many
855 client connections are expected to not authenticate in this way, so
856 an authenticate fail is not an error.
858 HS/02 Add the Exim version string to the process info. This way exiwhat
859 gives some more detail about the running daemon.
861 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
862 matter for fast-change records such as DNSBLs.
864 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
865 retry record, even if constant. There may be multiple transports with
866 different interface settings and the retry behaviour needs to be kept
869 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
871 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
873 JH/10 Bug 840: fix log_defer_output option of pipe transport
875 JH/11 Bug 830: use same host for all RCPTS of a message, even under
876 hosts_randomize. This matters a lot when combined with mua_wrapper.
878 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
879 ${quote_pgsql:<string>} operator.
881 JH/13 Bug 1708: avoid misaligned access in cached lookup.
883 JH/14 Change header file name for freeradius-client. Relevant if compiling
884 with Radius support; from the Gentoo tree and checked under Fedora.
886 JH/15 Bug 1712: Introduce $prdr_requested flag variable
888 JH/16 Bug 1714: Permit an empty string as expansion result for transport
889 option transport_filter, meaning no filtering.
891 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
893 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
894 defaults to "*" (all hosts). The variable is now available when not built
895 with TLS, default unset, mainly to enable keeping the testsuite sane.
896 If a server certificate is not supplied (via tls_certificate) an error is
897 logged, and clients will find TLS connections fail on startup. Presumably
898 they will retry in-clear.
899 Packagers of Exim are strongly encouraged to create a server certificate
900 at installation time.
902 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
903 with the $config_file variable.
905 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
906 in transport context, after the attempt, and per-recipient. The latter type
907 is per host attempted. The event data is the error message, and the errno
908 information encodes the lookup type (A vs. MX) used for the (first) host,
909 and the trailing two digits of the smtp 4xx response.
911 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
912 to write to mainlog (or rejectlog, paniclog) in the window between file
913 creation and permissions/ownership being changed. Particularly affects
914 installations where exicyclog is run as root, rather than exim user;
915 result is that the running daemon panics and dies.
917 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
919 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
920 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
921 "pri" and "weight". Note that the previous implicit priority given by the
922 list order is no longer honoured.
924 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
927 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
928 by defining SUPPORT_SOCKS.
930 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
931 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
932 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
933 variables are renamed to proxy_{local,external}_{address,port}.
935 JH/27 Move Internationalisation support from Experimental to mainline, enabled
936 for a build by defining SUPPORT_I18N
938 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
939 of the query string, and make ${quote_redis:} do that quoting.
941 JH/29 Move Events support from Experimental to mainline, enabled by default
942 and removable for a build by defining DISABLE_EVENT.
944 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
946 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
947 cached by the daemon.
949 JH/32 Move Redis support from Experimental to mainline, enabled for a build
950 by defining LOOKUP_REDIS. The libhiredis library is required.
952 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
953 keys are given for lookup.
955 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
956 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
957 only supported when built with TLS support. The PolarSSL SHA routines
958 are still used when the TLS library is too old for convenient support.
960 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
961 openssl_options), for security. OpenSSL forces this from version 1.1.0
962 server-side so match that on older versions.
964 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
965 allocation for $value could be released as the expansion processing
966 concluded, but leaving the global pointer active for it.
968 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
969 and to use the domains and local_parts ACL conditions.
971 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
972 incorrectly not doubled on cutthrough transmission, hence seen as a
973 body-termination at the receiving system - resulting in truncated mails.
974 Commonly the sender saw a TCP-level error, and retransmitted the message
975 via the normal store-and-forward channel. This could result in duplicates
976 received - but deduplicating mailstores were liable to retain only the
977 initial truncated version.
979 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
981 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
983 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
984 we're in there, support oversigning also; bug 1309.
986 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
988 HS/04 Add support for keep_environment and add_environment options.
990 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
991 either intentional arithmetic overflow during PRNG, or testing config-
994 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
995 delivery resulted in actual delivery. Cancel cutthrough before DATA
998 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
999 rejects a recipient: pass the reject to the originator.
1001 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1002 Many were false-positives and ignorable, but it's worth fixing the
1005 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1006 for the new environment-manipulation done at startup. Move the routines
1007 from being local to tls.c to being global via the os.c file.
1009 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1010 an extract embedded as result-arg for a map, the first arg for extract
1011 is unavailable so we cannot tell if this is a numbered or keyed
1012 extraction. Accept either.
1018 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1021 JH/02 The smtp transport option "multi_domain" is now expanded.
1023 JH/03 The smtp transport now requests PRDR by default, if the server offers
1026 JH/04 Certificate name checking on server certificates, when exim is a client,
1027 is now done by default. The transport option tls_verify_cert_hostnames
1028 can be used to disable this per-host. The build option
1029 EXPERIMENTAL_CERTNAMES is withdrawn.
1031 JH/05 The value of the tls_verify_certificates smtp transport and main options
1032 default to the word "system" to access the system default CA bundle.
1033 For GnuTLS, only version 3.0.20 or later.
1035 JH/06 Verification of the server certificate for a TLS connection is now tried
1036 (but not required) by default. The verification status is now logged by
1037 default, for both outbound TLS and client-certificate supplying inbound
1040 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1043 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1044 Status Notification (bounce) messages are now MIME format per RFC 3464.
1045 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1046 under the control of the dsn_advertise_hosts option, and routers may
1047 have a dsn_lasthop option.
1049 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1050 default, modifiable by a malware= option. The list separator for
1051 the options can now be changed in the usual way. Bug 68.
1053 JH/10 The smtp_receive_timeout main option is now expanded before use.
1055 JH/11 The incoming_interface log option now also enables logging of the
1056 local interface on delivery outgoing connections.
1058 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1059 if the interface and destination host and port all match.
1061 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1064 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1065 Patch from Andrew Lewis.
1067 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1068 now supports optional time-restrictions, weighting, and priority
1069 modifiers per server. Patch originally by <rommer@active.by>.
1071 JH/16 The spamd_address main option now supports a mixed list of local
1072 and remote servers. Remote servers can be IPv6 addresses, and
1073 specify a port-range.
1075 JH/17 Bug 68: The spamd_address main option now supports an optional
1076 timeout value per server.
1078 JH/18 Bug 1581: Router and transport options headers_add/remove can
1079 now have the list separator specified.
1081 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1084 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1087 JH/21 Support for the A6 type of dns record is withdrawn.
1089 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1090 rather than the verbs used.
1092 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1093 from 255 to 1024 chars.
1095 JH/24 Verification callouts now attempt to use TLS by default.
1097 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1098 are generic router options now. The defaults didn't change.
1100 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1101 Original patch from Alexander Shikoff, worked over by JH.
1103 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1104 files if ZCAT_COMMAND is not executable.
1106 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1108 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1110 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1111 Normally benign, it bites when the pair was led to by a CNAME;
1112 modern usage is to not canonicalize the domain to a CNAME target
1113 (and we were inconsistent anyway for A-only vs AAAA+A).
1115 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1117 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1118 when evaluating $sender_host_dnssec.
1120 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1121 $sender_helo_dnssec variable.
1123 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1125 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1127 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1129 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1130 documented as working, but never had. Support all but $spam_report.
1132 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1133 added for tls authenticator.
1135 HS/03 Add perl_taintmode main config option
1141 TL/01 When running the test suite, the README says that variables such as
1142 no_msglog_check are global and can be placed anywhere in a specific
1143 test's script, however it was observed that placement needed to be near
1144 the beginning for it to behave that way. Changed the runtest perl
1145 script to read through the entire script once to detect and set these
1146 variables, reset to the beginning of the script, and then run through
1147 the script parsing/test process like normal.
1149 TL/02 The BSD's have an arc4random API. One of the functions to induce
1150 adding randomness was arc4random_stir(), but it has been removed in
1151 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1152 function when detected.
1154 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1155 cause callback expansion.
1157 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1158 syntax errors in an expansion can be treated as a string instead of
1159 logging or causing an error, due to the internal use of bool_lax
1160 instead of bool when processing it.
1162 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1163 server certificates when making smtp deliveries.
1165 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1167 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1169 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1171 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1172 Merged patch from Sebastian Wiedenroth.
1174 JH/05 Fix results-pipe from transport process. Several recipients, combined
1175 with certificate use, exposed issues where response data items split
1176 over buffer boundaries were not parsed properly. This eventually
1177 resulted in duplicates being sent. This issue only became common enough
1178 to notice due to the introduction of connection certificate information,
1179 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1181 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1182 size buffer was used, resulting in syntax errors when an expansion
1185 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1186 version 3.3.6 or later.
1188 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1189 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1190 both become "event_action", the variables become $event_name, $event_data
1191 and $event_defer_errno. There is a new variable $verify_mode, usable in
1192 routers, transports and related events. The tls:cert event is now also
1193 raised for inbound connections, if the main configuration event_action
1196 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1197 early OCSP support, but no stapling (appears to be less than 1.0.0).
1199 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1200 server certificate names available under the smtp transport option
1201 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1204 JH/10 Time-related extraction expansions from certificates now use the main
1205 option "timezone" setting for output formatting, and are consistent
1206 between OpenSSL and GnuTLS compilations. Bug 1541.
1208 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1209 encoded parameter in the incoming message. Bug 1558.
1211 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1212 include certificate info, eximon was claiming there were spoolfile
1215 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1217 JH/14 Log delivery-related information more consistently, using the sequence
1218 "H=<name> [<ip>]" wherever possible.
1220 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1221 are problematic for Debian distribution, omit them from the release
1224 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1226 JH/16 Fix string representation of time values on 64bit time_t architectures.
1229 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1230 output list separator was used.
1235 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1236 checkers that were complaining about end of non-void function with no
1239 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1240 This was a regression introduced in 4.83 by another bugfix.
1242 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1244 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1245 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1251 TF/01 Correctly close the server side of TLS when forking for delivery.
1253 When a message was received over SMTP with TLS, Exim failed to clear up
1254 the incoming connection properly after forking off the child process to
1255 deliver the message. In some situations the subsequent outgoing
1256 delivery connection happened to have the same fd number as the incoming
1257 connection previously had. Exim would try to use TLS and fail, logging
1258 a "Bad file descriptor" error.
1260 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1261 utilities have not been installed.
1263 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1264 temporary space as the ACL may create new global variables.
1266 TL/01 LDAP support uses per connection or global context settings, depending
1267 upon the detected version of the libraries at build time.
1269 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1270 to extract and use the src ip:port in logging and expansions as if it
1271 were a direct connection from the outside internet. PPv2 support was
1272 updated based on HAProxy spec change in May 2014.
1274 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1276 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1277 Properly escape header and check for NULL return.
1279 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1282 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1284 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1285 characters in header names, implemented as a verify condition.
1286 Contributed by Michael Fischer v. Mollard.
1288 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1289 results permerror and temperror. Previous values are deprecated but
1290 still accepted. In a future release, err_perm and err_temp will be
1291 completely removed, which will be a backward incompatibility if the
1292 ACL tests for either of these two old results. Patch contributed by
1293 user bes-internal on the mailing list.
1295 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1297 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1298 selectors, in both main and reject logs.
1300 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1303 JH/07 Add malware type "sock" for talking to simple daemon.
1305 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1307 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1308 routers/transports under cutthrough routing.
1310 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1311 numbers. Touch up "bool" conditional to keep the same definition.
1313 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1315 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1317 JH/12 Expand items in router/transport headers_add or headers_remove lists
1318 individually rather than the list as a whole. Bug 1452.
1320 Required for reasonable handling of multiple headers_ options when
1321 they may be empty; requires that headers_remove items with embedded
1322 colons must have them doubled (or the list-separator changed).
1324 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1325 view the policy declared in the DMARC record. Currently, $dmarc_status
1326 is a combined value of both the record presence and the result of the
1329 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1331 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1332 dnslookup router and the smtp transport (applying to the forward
1335 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1336 of ldap servers used for a specific lookup. Patch provided by Heiko
1339 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1340 New variable $lookup_dnssec_authenticated for observability.
1342 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1343 Patch submitted by Lars Timman.
1345 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1347 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1348 Requires trusted mode and valid format message id, aborts otherwise.
1349 Patch contributed by Heiko Schlichting.
1351 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1352 certextract with support for various fields. Bug 1358.
1354 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1355 is requested by default, modifiable by smtp transport option
1358 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1359 operate on certificate variables to give certificate fingerprints
1360 Also new ${sha256:cert_variable}.
1362 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1364 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1365 Christian Aistleitner.
1367 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1369 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1370 file. Patch from Wolfgang Breyha.
1372 JH/25 Expand the coverage of the delivery $host and $host_address to
1373 client authenticators run in verify callout. Bug 1476.
1375 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1376 align with daemon_smtp_ports. Bug 72.
1378 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1379 support and error reporting did not work properly.
1381 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1382 and is readable. Patch from Andrew Colin Kissa.
1384 TL/14 Enhance documentation of ${run expansion and how it parses the
1385 commandline after expansion, particularly in the case when an
1386 unquoted variable expansion results in an empty value.
1388 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1390 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1391 by renaming away from it. Addresses GH issue 15, reported by
1394 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1396 TL/15 SECURITY: prevent double expansion in math comparison functions
1397 (can expand unsanitized data). Not remotely exploitable.
1404 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1406 PP/02 Make -n do something, by making it not do something.
1407 When combined with -bP, the name of an option is not output.
1409 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1412 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1413 $sender_host_name and config options to manage this, and basic check
1416 PP/05 DSCP support for outbound connections and control modifier for inbound.
1418 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1419 (Only plugin which currently uses this is kerberos4, which nobody should
1420 be using, but we should make it available and other future plugins might
1421 conceivably use it, even though it would break NAT; stuff *should* be
1422 using channel bindings instead).
1424 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1425 name; added for Sendmail compatibility; requires admin caller.
1426 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1427 just ignore it); requires trusted caller.
1428 Also parse but ignore: -Ac -Am -X<logfile>
1431 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1433 TL/02 Add +smtp_confirmation as a default logging option.
1435 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1436 Patch by Magnus Holmgren from 2007-02-20.
1438 TL/04 Bugzilla 1281 - Spec typo.
1439 Bugzilla 1283 - Spec typo.
1440 Bugzilla 1290 - Spec grammar fixes.
1442 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1444 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1446 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1447 mailing list by by Dmitry Isaikin.
1449 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1451 JH/02 Support "G" suffix to numbers in ${if comparisons.
1453 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1455 NM/01 Bugzilla 1197 - Spec typo
1456 Bugzilla 1196 - Spec examples corrections
1458 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1460 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1461 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1464 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1465 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1467 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1468 "acl {{name}{arg}...}", and optional args on acl condition
1471 JH/05 Permit multiple router/transport headers_add/remove lines.
1473 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1475 JH/07 Avoid using a waiting database for a single-message-only transport.
1476 Performance patch from Paul Fisher. Bugzilla 1262.
1478 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1481 JH/09 Add $headers_added variable, with content from use of ACL modifier
1482 add_header (but not yet added to the message). Bugzilla 199.
1484 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1485 Pulled from Bugzilla 817 by Wolfgang Breyha.
1487 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1489 (nb: this is the same fix as in Exim 4.80.1)
1491 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1494 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1497 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1499 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1500 advertise SMTP AUTH mechanism to us, instead of a generic
1501 protocol violation error. Also, make Exim more robust to bad
1502 data from the Dovecot auth socket.
1504 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1506 When a queue runner is handling a message, Exim first routes the
1507 recipient addresses, during which it prunes them based on the retry
1508 hints database. After that it attempts to deliver the message to
1509 any remaining recipients. It then updates the hints database using
1512 So if a recipient address works intermittently, it can get repeatedly
1513 deferred at routing time. The retry hints record remains fresh so the
1514 address never reaches the final cutoff time.
1516 This is a fairly common occurrence when a user is bumping up against
1517 their storage quota. Exim had some logic in its local delivery code
1518 to deal with this. However it did not apply to per-recipient defers
1519 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1521 This change adds a proper retry rule check during routing so that the
1522 final cutoff time is checked against the message's age. We only do
1523 this check if there is an address retry record and there is not a
1524 domain retry record; this implies that previous attempts to handle
1525 the address had the retry_use_local_parts option turned on. We use
1526 this as an approximation for the destination being like a local
1527 delivery, as in LMTP.
1529 I suspect this new check makes the old local delivery cutoff check
1530 redundant, but I have not verified this so I left the code in place.
1532 TF/02 Correct gecos expansion when From: is a prefix of the username.
1534 Test 0254 submits a message to Exim with the header
1538 When I ran the test suite under the user fanf2, Exim expanded
1539 the header to contain my full name, whereas it should have added
1540 a Resent-Sender: header. It erroneously treats any prefix of the
1541 username as equal to the username.
1543 This change corrects that bug.
1545 GF/01 DCC debug and logging tidyup
1546 Error conditions log to paniclog rather than rejectlog.
1547 Debug lines prefixed by "DCC: " to remove any ambiguity.
1549 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1551 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1552 Bug spotted by Jeremy Harris; was flawed since initial commit.
1553 Would have resulted in OCSP responses post-SNI triggering an Exim
1554 NULL dereference and crash.
1556 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1558 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1559 Bug detection, analysis and fix by Samuel Thibault.
1560 Bugzilla 1331, Debian bug #698092.
1562 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1564 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1565 Server implementation by Todd Lyons, client by JH.
1566 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1567 config variable "prdr_enable" controls whether the server
1568 advertises the facility. If the client requests PRDR a new
1569 acl_data_smtp_prdr ACL is called once for each recipient, after
1570 the body content is received and before the acl_smtp_data ACL.
1571 The client is controlled by both of: a hosts_try_prdr option
1572 on the smtp transport, and the server advertisement.
1573 Default client logging of deliveries and rejections involving
1574 PRDR are flagged with the string "PRDR".
1576 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1577 fclose(). Diagnosis by Todd Lyons.
1579 PP/17 Update configure.default to handle IPv6 localhost better.
1580 Patch by Alain Williams (plus minor tweaks).
1583 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1584 This is now consistent with GnuTLS, and is now documented: the
1585 previous undocumented portable approach to treating the option as
1586 unset was to force an expansion failure. That still works, and
1587 an empty string is now equivalent.
1589 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1590 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1591 not performing validation itself.
1593 PP/20 Added force_command boolean option to pipe transport.
1594 Patch from Nick Koston, of cPanel Inc.
1596 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1599 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1601 PP/21 Fix eximon continuous updating with timestamped log-files.
1602 Broken in a format-string cleanup in 4.80, missed when I repaired the
1603 other false fix of the same issue.
1604 Report and fix from Heiko Schlichting.
1607 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1608 Report from Prashanth Katuri.
1610 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1611 It's SecureTransport, so affects any MacOS clients which use the
1612 system-integrated TLS libraries, including email clients.
1614 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1615 using a MIME ACL for non-SMTP local injection.
1616 Report and assistance in diagnosis by Warren Baker.
1618 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1620 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1622 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1623 last id that failed so it may be referenced in subsequent ACL's.
1625 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1628 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1629 ldap library initialization, allowing self-signed CA's to be
1630 used. Also properly sets require_cert option later in code by
1631 using NULL (global ldap config) instead of ldap handle (per
1632 session). Bug diagnosis and testing by alxgomz.
1634 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1635 the src/util/ subdirectory.
1637 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1638 renamed to Transport Post Delivery Action by Jeremy Harris, as
1641 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1642 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1643 redis_servers = needs to be configured which will be used by the redis
1644 lookup. Patch from Warren Baker, of The Packet Hub.
1646 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1648 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1649 hostname or reverse DNS when processing a host list. Used suggestions
1650 from multiple comments on this bug.
1652 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1654 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1655 Missed a few lines, added it to make the runtest require no keyboard
1658 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1659 contains upper case chars. Make router use caseful_local_part.
1661 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1662 support when GnuTLS has been built with p11-kit.
1668 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1670 This, or similar/improved, will also be change PP/11 of 4.82.
1676 PP/01 Handle short writes when writing local log-files.
1677 In practice, only affects FreeBSD (8 onwards).
1678 Bugzilla 1053, with thanks to Dmitry Isaikin.
1680 NM/01 Bugzilla 949 - Documentation tweak
1682 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1685 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1687 PP/02 Implemented gsasl authenticator.
1689 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1691 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1692 `pkg-config foo` for cflags/libs.
1694 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1695 with rest of GSASL and with heimdal_gssapi.
1697 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1698 `pkg-config foo` for cflags/libs for the TLS implementation.
1700 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1701 properties get this fed in as external SSF. A number of robustness
1702 and debugging improvements to the cyrus_sasl authenticator.
1704 PP/08 cyrus_sasl server now expands the server_realm option.
1706 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1707 Patch by Jeremy Harris.
1709 PP/10 Added dbmjz lookup type.
1711 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1713 PP/12 MAIL args handles TAB as well as SP, for better interop with
1714 non-compliant senders.
1715 Analysis and variant patch by Todd Lyons.
1717 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1718 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1719 Patch from Dirk Mueller <dmueller@suse.com>
1721 PP/13 tls_peerdn now print-escaped for spool files.
1722 Observed some $tls_peerdn in wild which contained \n, which resulted
1723 in spool file corruption.
1725 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1726 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1727 or write after TLS renegotiation, which otherwise led to messages
1730 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1731 as a tracking header (ie: a signed header comes before the signature).
1732 Patch from Wolfgang Breyha.
1734 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1735 comma-sep list; embedded commas doubled.
1737 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1739 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1741 Report and patch from Dmitry Banschikov.
1743 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1744 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1745 failures. We appear to now support TLS1.1+ with Exim.
1747 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1748 lets Exim select keys and certificates based upon TLS SNI from client.
1749 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1750 before an outbound SMTP session. New log_selector, +tls_sni.
1752 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1753 NULL dereference. Report and patch from Alun Jones.
1755 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1756 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1757 Not seeing resolver debug output on NetBSD, but suspect this is a
1758 resolver implementation change.
1760 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1761 Left warnings. Added "eximon gdb" invocation mode.
1763 PP/21 Defaulting "accept_8bitmime" to true, not false.
1765 PP/22 Added -bw for inetd wait mode support.
1767 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1768 locate the relevant includes and libraries. Made this the default.
1770 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1771 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1773 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1774 This may cause build issues on older platforms.
1776 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1777 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1778 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1779 Added SNI support via GnuTLS too.
1780 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1782 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1784 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1785 Applied second patch from Janne, implementing suggestion to default
1786 multiple-strings-in-record handling to match SPF spec.
1788 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1790 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1791 read-only, out of scope).
1792 Patch from Wolfgang Breyha, report from Stuart Northfield.
1794 PP/29 Fix three issues highlighted by clang analyser static analysis.
1795 Only crash-plausible issue would require the Cambridge-specific
1796 iplookup router and a misconfiguration.
1797 Report from Marcin Mirosław.
1799 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1801 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1802 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1803 As part of this, removing so much warning spew let me fix some minor
1804 real issues in debug logging.
1806 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1807 assignment on my part. Fixed.
1809 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1810 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1811 Janne Snabb (who went above and beyond: thank you).
1813 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1814 string otherwise requires a connection and a bunch more work and it's
1815 relatively easy to get wrong. Should also expose TLS library linkage
1818 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1819 64-bit ${eval} (JH/03).
1821 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1822 GNU libc to support some of the 64-bit stuff, should not lead to
1823 conflicts. Defined before os.h is pulled in, so if a given platform
1824 needs to override this, it can.
1826 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1827 protection layer was required, which is not implemented.
1828 Bugzilla 1254, patch from Wolfgang Breyha.
1830 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1831 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1832 tls_dhparam take prime identifiers. Also unbreak combination of
1833 OpenSSL+DH_params+TLSSNI.
1835 PP/39 Disable SSLv2 by default in OpenSSL support.
1841 PP/01 Solaris build fix for Oracle's LDAP libraries.
1842 Bugzilla 1109, patch from Stephen Usher.
1844 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1846 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1849 TF/02 Fix a couple more cases where we did not log the error message
1850 when unlink() failed. See also change 4.74-TF/03.
1852 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1853 lock up or crash if it happened to be inside a call to libc when it
1854 got a SIGUSR1 from exiwhat.
1856 The SIGUSR1 handler appends the current process status to the process
1857 log which is later printed by exiwhat. It used to use the general
1858 purpose logging code to do this, but several functions it calls are
1859 not safe for signals.
1861 The new output code in the SIGUSR1 handler is specific to the process
1862 log, and simple enough that it's easy to inspect for signal safety.
1863 Removing some special cases also simplifies the general logging code.
1864 Removing the spurious timestamps from the process log simplifies
1867 TF/04 Improved ratelimit ACL condition.
1869 The /noupdate option has been deprecated in favour of /readonly which
1870 has clearer semantics. The /leaky, /strict, and /readonly update modes
1871 are mutually exclusive. The update mode is no longer included in the
1872 database key; it just determines when the database is updated. (This
1873 means that when you upgrade Exim will forget old rate measurements.)
1875 Exim now checks that the per_* options are used with an update mode that
1876 makes sense for the current ACL. For example, when Exim is processing a
1877 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1878 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1879 must specify per_mail/readonly. If you omit the update mode it defaults to
1880 /leaky where that makes sense (as before) or /readonly where required.
1882 The /noupdate option is now undocumented but still supported for
1883 backwards compatibility. It is equivalent to /readonly except that in
1884 ACLs where /readonly is required you may specify /leaky/noupdate or
1885 /strict/noupdate which are treated the same as /readonly.
1887 A useful new feature is the /count= option. This is a generalization
1888 of the per_byte option, so that you can measure the throughput of other
1889 aggregate values. For example, the per_byte option is now equivalent
1890 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1892 The per_rcpt option has been generalized using the /count= mechanism
1893 (though it's more complicated than the per_byte equivalence). When it is
1894 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1895 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1896 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1897 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1898 Note that using per_rcpt with a non-readonly update mode in more than
1899 one ACL will cause the recipients to be double-counted. (The per_mail
1900 and per_byte options don't have this problem.)
1902 The handling of very low rates has changed slightly. If the computed rate
1903 is less than the event's count (usually one) then this event is the first
1904 after a long gap. In this case the rate is set to the same as this event's
1905 count, so that the first message of a spam run is counted properly.
1907 The major new feature is a mechanism for counting the rate of unique
1908 events. The new per_addr option counts the number of different
1909 recipients that someone has sent messages to in the last time period. It
1910 behaves like per_rcpt if all the recipient addresses are different, but
1911 duplicate recipient addresses do not increase the measured rate. Like
1912 the /count= option this is a general mechanism, so the per_addr option
1913 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1914 example, measure the rate that a client uses different sender addresses
1915 with the options per_mail/unique=$sender_address. There are further
1916 details in the main documentation.
1918 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1920 TF/06 Removed a few PCRE remnants.
1922 TF/07 Automatically extract Exim's version number from tags in the git
1923 repository when doing development or release builds.
1925 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1926 Bugzilla 879. Patch from Paul Fisher.
1928 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1929 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1932 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1934 PP/05 Variable $av_failed, true if the AV scanner deferred.
1935 Bugzilla 1078. Patch from John Horne.
1937 PP/06 Stop make process more reliably on build failure.
1938 Bugzilla 1087. Patch from Heiko Schlittermann.
1940 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1941 Bugzilla 1089. Patch from Heiko Schlittermann.
1943 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1944 Bugzilla 1131. Patch from Holger Weiß.
1946 PP/09 Handle IPv6 addresses with SPF.
1947 Bugzilla 860. Patch from Wolfgang Breyha.
1949 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1951 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1954 PP/11 match_* no longer expand right-hand-side by default.
1955 New compile-time build option, EXPAND_LISTMATCH_RHS.
1956 New expansion conditions, "inlist", "inlisti".
1958 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1960 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1962 PP/14 fix log_write() format string regression from TF/03.
1963 Bugzilla 1152. Patch from Dmitry Isaikin.
1969 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1971 PP/02 Harmonised TLS library version reporting; only show if debugging.
1972 Layout now matches that introduced for other libraries in 4.74 PP/03.
1974 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1976 PP/04 New "dns_use_edns0" global option.
1978 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1981 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1982 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1984 TK/01 Updated PolarSSL code to 0.14.2.
1985 Bugzilla 1097. Patch from Andreas Metzler.
1987 PP/07 Catch divide-by-zero in ${eval:...}.
1988 Fixes bugzilla 1102.
1990 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1993 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1994 format-string attack -- SECURITY: remote arbitrary code execution.
1996 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1997 time unintentionally subject to list matching rules, letting the header
1998 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1999 arbitrary string expansion). This allowed for information disclosure.
2001 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2002 INT_MIN/-1 -- value coerced to INT_MAX.
2008 NM/01 Workaround for PCRE version dependency in version reporting
2011 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2012 This fixes portability to compilers other than gcc, notably
2013 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2015 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2016 makefiles for portability to HP-UX and POSIX correctness.
2018 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2019 Also via indented variable definition in the Makefile.
2020 (Debugging by Oliver Heesakkers).
2022 PP/02 Restore caching of spamd results with expanded spamd_address.
2023 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2025 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2026 Improves build reliability. Fix from: Frank Elsner
2028 NM/02 Fix wide character breakage in the rfc2047 coding
2029 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2031 NM/03 Allow underscore in dnslist lookups
2032 Fixes bug 1026. Patch from Graeme Fowler
2034 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2035 Code patches from Adam Ciarcinski of NetBSD.
2037 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2040 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2041 is logged, to avoid truncation. Patch from John Horne.
2043 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2044 Patch from Jakob Hirsch.
2046 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2047 SQL string expansion failure details.
2048 Patch from Andrey Oktyabrski.
2050 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2051 Patch from Simon Arlott.
2053 PP/09 New lookups functionality failed to compile on old gcc which rejects
2054 extern declarations in function scope.
2055 Patch from Oliver Fleischmann
2057 PP/10 Use sig_atomic_t for flags set from signal handlers.
2058 Check getgroups() return and improve debugging.
2059 Fixed developed for diagnosis in bug 927 (which turned out to be
2062 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2063 Patch from Mark Zealey.
2065 PP/12 Bugzilla 1056: Improved spamd server selection.
2066 Patch from Mark Zealey.
2068 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2069 Based on patch from Heiko Schlittermann.
2071 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2072 Patch from Uwe Doering, sign-off by Michael Haardt.
2074 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2075 variable declaration deep within a block. Bug and patch from
2078 PP/15 lookups-Makefile IRIX compatibility coercion.
2080 PP/16 Make DISABLE_DKIM build knob functional.
2082 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2083 Patch by Simon Arlott
2085 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2086 variable argument macros. Our copy now differs from upstream.
2092 TF/01 Failure to get a lock on a hints database can have serious
2093 consequences so log it to the panic log.
2095 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2096 controlled using the smtp_confirmation log selector.
2098 TF/03 Include the error message when we fail to unlink a spool file.
2100 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2101 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2102 for maintaining out-of-tree patches for some time.
2104 PP/01 Bugzilla 139: Documentation and portability issues.
2105 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2106 Handle per-OS dynamic-module compilation flags.
2108 PP/02 Let /dev/null have normal permissions.
2109 The 4.73 fixes were a little too stringent and complained about the
2110 permissions on /dev/null. Exempt it from some checks.
2111 Reported by Andreas M. Kirchwitz.
2113 PP/03 Report version information for many libraries, including
2114 Exim version information for dynamically loaded libraries. Created
2115 version.h, now support a version extension string for distributors
2116 who patch heavily. Dynamic module ABI change.
2118 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2119 privilege escalation vulnerability whereby the Exim run-time user
2120 can cause root to append content of the attacker's choosing to
2123 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2126 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2127 If dropping privileges for untrusted macros, we disabled normal logging
2128 on the basis that it would fail; for the Exim run-time user, this is not
2129 the case, and it resulted in successful deliveries going unlogged.
2130 Fixed. Reported by Andreas Metzler.
2136 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2137 only prepend for the Resent-* case. Fixes regression introduced in
2138 Exim 4.70 by NM/22 for Bugzilla 607.
2140 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2141 increasing numbers of administrators be bitten by this.
2143 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2145 PP/03 Bugzilla 994: added openssl_options main configuration option.
2147 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2149 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2151 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2153 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2154 without a peer certificate, leading to a segfault because of an
2155 assumption that peers always have certificates. Be a little more
2156 paranoid. Problem reported by Martin Tscholak.
2158 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2159 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2160 NB: ClamAV planning to remove STREAM in "middle of 2010".
2161 CL also introduces -bmalware, various -d+acl logging additions and
2162 more caution in buffer sizes.
2164 PP/09 Implemented reverse_ip expansion operator.
2166 PP/10 Bugzilla 937: provide a "debug" ACL control.
2168 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2170 PP/12 Bugzilla 973: Implement --version.
2172 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2174 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2176 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2178 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2179 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2180 ignore trailing whitespace.
2182 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2184 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2187 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2188 Notification from Dr Andrew Aitchison.
2190 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2191 ExtendedDetectionInfo response format.
2192 Notification from John Horne.
2194 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2197 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2198 XSL and documented dependency on system catalogs, with examples of how
2201 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2204 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2205 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2206 configuration file which is writeable by the Exim user or group.
2208 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2209 of configuration files to cover files specified with the -C option if
2210 they are going to be used with root privileges, not just the default
2213 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2214 option (effectively making it always true).
2216 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2217 files to be used while preserving root privileges.
2219 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2220 that rogue child processes cannot use them.
2222 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2223 run-time user, instead of root.
2225 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2226 Exim run-time user without dropping privileges.
2228 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2229 result string, instead of calling string_vformat() twice with the same
2232 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2233 for other users. Others should always drop root privileges if they use
2234 -C on the command line, even for a whitelisted configure file.
2236 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2238 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2244 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2245 $data_path, and $header_path variables; fixed documentation bugs and
2248 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2249 exipick to access non-standard spools, including the "frozen" queue
2252 NM/01 Bugzilla 965: Support mysql stored procedures.
2253 Patch from Alain Williams
2255 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2257 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2258 Patch from Andreas Metzler
2260 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2261 Patch from Kirill Miazine
2263 NM/05 Bugzilla 671: Added umask to procmail example.
2265 JJ/03 installed exipick 20100323.0, fixing doc bug
2267 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2268 directory. Notification and patch from Dan Rosenberg.
2270 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2272 TK/02 Improve log output when DKIM signing operation fails.
2274 MH/01 Treat the transport option dkim_domain as a colon separated
2275 list, not as a single string, and sign the message with each element,
2276 omitting multiple occurences of the same signer.
2278 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2279 Bugzilla 985, 986. Patch by Simon Arlott
2281 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2282 Patch by Simon Arlott
2284 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2285 MBX locking. Notification from Dan Rosenberg.
2291 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2293 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2295 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2297 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2299 NM/04 Bugzilla 915: Flag broken perl installation during build.
2305 TK/01 Added patch by Johannes Berg that expands the main option
2306 "spamd_address" if it starts with a dollar sign.
2308 TK/02 Write list of recipients to X-Envelope-Sender header when building
2309 the mbox-format spool file for content scanning (suggested by Jakob
2312 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2313 (http://www.dcc-servers.net/) support via dccifd. Activated by
2314 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2316 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2317 by Mark Daniel Reidel <mr@df.eu>.
2319 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2320 When building exim an external PCRE library is now needed -
2321 PCRE is a system library on the majority of modern systems.
2322 See entry on PCRE_LIBS in EDITME file.
2324 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2325 conversation. Added nologin parameter to request.
2326 Patch contributed by Kirill Miazine.
2328 TF/01 Do not log submission mode rewrites if they do not change the address.
2330 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2332 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2333 log files in place. Contributed by Roberto Lima.
2335 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2337 TF/03 Bugzilla 615: When checking the local_parts router precondition
2338 after a local_part_suffix or local_part_prefix option, Exim now
2339 does not use the address's named list lookup cache, since this
2340 contains cached lookups for the whole local part.
2342 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2343 Robert Millan. Documentation is in experimental-spec.txt.
2345 TF/04 Bugzilla 668: Fix parallel build (make -j).
2347 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2349 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2350 Patch provided by Jan Srzednicki.
2352 TF/05 Leading white space used to be stripped from $spam_report which
2353 wrecked the formatting. Now it is preserved.
2355 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2356 that they are available at delivery time.
2358 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2360 TF/08 TLS error reporting now respects the incoming_interface and
2361 incoming_port log selectors.
2363 TF/09 Produce a more useful error message if an SMTP transport's hosts
2364 setting expands to an empty string.
2366 NM/06 Bugzilla 744: EXPN did not work under TLS.
2367 Patch provided by Phil Pennock.
2369 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2370 Patch provided by Richard Godbee.
2372 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2373 acl_smtp_notquit, added index entry.
2375 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2376 Patch provided by Eugene Bujak.
2378 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2379 accept(). Patch provided by Maxim Dounin.
2381 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2382 Patch provided by Phil Pennock.
2384 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2386 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2387 Patch provided by Brad "anomie" Jorsch.
2389 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2390 Patch provided by Dean Brooks.
2392 TK/05 Add native DKIM support (does not depend on external libraries).
2394 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2395 Patch provided by Graeme Fowler.
2397 NM/16 Bugzilla 851: Documentation example syntax fix.
2399 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2401 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2404 NM/19 Bugzilla 745: TLS version reporting.
2405 Patch provided by Phil Pennock.
2407 NM/20 Bugzilla 167: bool: condition support.
2408 Patch provided by Phil Pennock.
2410 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2411 clients. Patch provided by Phil Pennock.
2413 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2414 Patch provided by Brad "anomie" Jorsch.
2416 NM/23 Bugzilla 687: Fix misparses in eximstats.
2417 Patch provided by Heiko Schlittermann.
2419 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2420 Patch provided by Heiko Schlittermann.
2422 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2423 plus update to original patch.
2425 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2427 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2428 Patch provided by David Brownlee.
2430 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2432 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2434 NM/30 Bugzilla 888: TLS documentation bugfixes.
2436 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2438 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2439 Unlike the original bugzilla I have changed all shell scripts in src tree.
2441 NM/33 Bugzilla 898: Transport filter timeout fix.
2442 Patch by Todd Rinaldo.
2444 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2445 Patch by Serge Demonchaux.
2447 NM/35 Bugzilla 39: Base64 decode bug fixes.
2448 Patch by Jakob Hirsch.
2450 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2452 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2454 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2456 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2462 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2463 ALT-N's libdkim that I have put here:
2464 http://duncanthrax.net/exim-experimental/
2466 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2467 were called 'true' and it seems that C99 defines that as a reserved
2468 keyword to be used with 'bool' variable types. That means you could
2469 not include C99-style headers which use bools without triggering
2470 build errors in sieve.c.
2472 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2473 as mailq or other aliases. Changed the --help handling significantly
2474 to do whats expected. exim_usage() emits usage/help information.
2476 SC/01 Added the -bylocaldomain option to eximstats.
2478 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2480 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2482 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2488 PH/01 Another patch from the Sieve maintainer.
2490 PH/02 When an IPv6 address is converted to a string for single-key lookup
2491 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2492 dots are used instead of colons so that keys in lsearch files need not
2493 contain colons. This was done some time before quoting was made available
2494 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2495 (notated using the quote facility) so as to distinguish them from IPv4
2496 keys. This meant that lookups for IP addresses in host lists did not work
2497 for iplsearch lookups.
2499 This has been fixed by arranging for IPv6 addresses to be expressed with
2500 colons if the lookup type is iplsearch. This is not incompatible, because
2501 previously such lookups could never work.
2503 The situation is now rather anomalous, since one *can* have colons in
2504 ordinary lsearch keys. However, making the change in all cases is
2505 incompatible and would probably break a number of configurations.
2507 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2510 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2511 conversion specification without a maximum field width, thereby enabling
2512 a rogue spamd server to cause a buffer overflow. While nobody in their
2513 right mind would setup Exim to query an untrusted spamd server, an
2514 attacker that gains access to a server running spamd could potentially
2515 exploit this vulnerability to run arbitrary code as the Exim user.
2517 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2518 $primary_hostname instead of what libspf2 thinks the hosts name is.
2520 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2521 a directory entry by the name of the lookup key. Previously, if a
2522 symlink pointed to a non-existing file or a file in a directory that
2523 Exim lacked permissions to read, a lookup for a key matching that
2524 symlink would fail. Now it is enough that a matching directory entry
2525 exists, symlink or not. (Bugzilla 503.)
2527 PH/03 The body_linecount and body_zerocount variables are now exported in the
2530 PH/04 Added the $dnslist_matched variable.
2532 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2533 This means they are set thereafter only if the connection becomes
2536 PH/06 Added the client_condition to authenticators so that some can be skipped
2537 by clients under certain conditions.
2539 PH/07 The error message for a badly-placed control=no_multiline_responses left
2540 "_responses" off the end of the name.
2542 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2544 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2545 (without spaces) instead of just copying the configuration text.
2547 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2549 PH/11 Added $max_received_linelength.
2551 PH/12 Added +ignore_defer and +include_defer to host lists.
2553 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2554 way in which PCRE > 7.0 is built.
2556 PH/14 Implemented queue_only_load_latch.
2558 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2559 MAIL command. The effect was to mangle the value on 64-bit systems.
2561 PH/16 Another patch from the Sieve maintainer.
2563 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2565 PH/18 If a system quota error occurred while trying to create the file for
2566 a maildir delivery, the message "Mailbox is full" was not appended to the
2567 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2568 only to a quota excession during the actual writing of the file.
2570 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2571 characters?) which causes problems in log lines. The DN values are now
2572 passed through string_printing() before being added to log lines.
2574 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2575 and InterBase are left for another time.)
2577 PH/21 Added message_body_newlines option.
2579 PH/22 Guard against possible overflow in moan_check_errorcopy().
2581 PH/23 POSIX allows open() to be a macro; guard against that.
2583 PH/24 If the recipient of an error message contained an @ in the local part
2584 (suitably quoted, of course), incorrect values were put in $domain and
2585 $local_part during the evaluation of errors_copy.
2591 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2592 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2595 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2596 issue a MAIL command.
2598 PH/02 In an ACL statement such as
2600 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2602 if a client was not listed at all, or was listed with a value other than
2603 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2604 the condition was not true (as it should be), so access was not denied.
2605 The bug was that the ! inversion was incorrectly passed on to the second
2606 item. This has been fixed.
2608 PH/03 Added additional dnslists conditions == and =& which are different from
2609 = and & when the dns lookup returns more than one IP address.
2611 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2612 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2614 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2615 FSYNC, which compiles an option called disable_fsync that allows for
2616 bypassing fsync(). The documentation is heavily laced with warnings.
2618 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2620 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2621 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2622 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2623 including adding "make clean"; (3) Added -fPIC when compiling the test
2624 dynamically loaded module, to get rid of a warning.
2626 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2627 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2628 = 0s. The bug is otherwise harmless.
2630 PH/07 There was a bug in the dovecot authenticator such that the value of
2631 $auth1 could be overwritten, and so not correctly preserved, after a
2632 successful authentication. This usually meant that the value preserved by
2633 the server_setid option was incorrect.
2635 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2637 PH/09 Installed PCRE release 7.0.
2639 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2640 run for batched SMTP input. It is now run at the start of every message
2641 in the batch. While fixing this I discovered that the process information
2642 (output by running exiwhat) was not always getting set for -bs and -bS
2643 input. This is fixed, and it now also says "batched" for BSMTP.
2645 PH/11 Added control=no_pipelining.
2647 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2648 patch, slightly modified), and move the expansion of helo_data till after
2649 the connection is made in the smtp transport (so it can use these
2652 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2654 PH/14 Added log_selector = +pid.
2656 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2658 PH/16 Add ${if forany and ${if forall.
2660 PH/17 Added dsn_from option to vary the From: line in DSNs.
2662 PH/18 Flush SMTP output before performing a callout, unless control =
2663 no_callout_flush is set.
2665 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2666 was true (the default) a successful delivery failed to delete the retry
2667 item, thus causing premature timeout of the address. The bug is now
2670 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2672 PH/21 Long custom messages for fakedefer and fakereject are now split up
2673 into multiline responses in the same way that messages for "deny" and
2674 other ACL rejections are.
2676 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2677 with slight modification.
2679 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2680 draft, changing the syntax and factoring some duplicate code".
2682 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2683 for deliveries of the second and subsequent messages over the same SMTP
2686 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2687 ${reduce, with only minor "tidies".
2689 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2691 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2692 expansion side effects.
2694 PH/27 When a message times out after an over-quota error from an Exim-imposed
2695 quota, the bounce message says "mailbox is full". This message was not
2696 being given when it was a system quota that was exceeded. It now should
2699 MH/03 Made $recipients available in local_scan(). local_scan() already has
2700 better access to the recipient list through recipients_list[], but
2701 $recipients can be useful in postmaster-provided expansion strings.
2703 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2704 in the case of a MAIL command with additional options following the
2705 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2706 were accidentally chopped off.
2708 PH/29 SMTP synchronization checks are implemented when a command is read -
2709 there is a check that no more input is waiting when there shouldn't be
2710 any. However, for some commands, a delay in an ACL can mean that it is
2711 some time before the response is written. In this time, more input might
2712 arrive, invalidly. So now there are extra checks after an ACL has run for
2713 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2714 pipelining has not been advertised.
2716 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2718 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2719 returned in the SMTP response when smtp_return_error_details was false.
2720 This has been fixed.
2722 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2723 instead of the C I/O that was originally supplied, because problems were
2724 reported on Solaris.
2726 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2727 Exim which did not show up earlier: it was assuming that a call to
2728 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2729 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2730 was a macro that became an assignment, so it seemed to work. This has
2731 changed to a proper function call with a void return, hence the compile
2732 error. Exim's code has been fixed.
2734 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2737 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2739 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2741 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2742 because it turns out to be too big or there is a timeout) there is no
2743 "Completed" line in the log. When some messages of this type were
2744 selected by exigrep, they were listed as "not completed". Others were
2745 picked up by some special patterns. I have improved the selection
2746 criteria to be more general.
2748 PH/38 The host_find_failed option in the manualroute router can now be set
2749 to "ignore", to completely ignore a host whose IP address cannot be
2750 found. If all hosts are ignored, the behaviour is controlled by the new
2751 host_all_ignored option.
2753 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2754 homing or because of multiple MX records with /mx) generated more than
2755 one IP address, and the following item turned out to be the local host,
2756 all the secondary addresses of the first item were incorrectly removed
2757 from the list, along with the local host and any following hosts (which
2758 is what is supposed to happen).
2760 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2761 whoever called Exim into the -H file. In the case of the daemon it was
2762 behaving confusingly. When first started, it used values for whoever
2763 started the daemon, but after a SIGHUP it used the Exim user (because it
2764 calls itself on a restart). I have changed the code so that it now always
2767 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2768 message are rejected with the same error (e.g. no authentication or bad
2769 sender address), and a DATA command is nevertheless sent (as can happen
2770 with PIPELINING or a stupid MUA), the error message that was given to the
2771 RCPT commands is included in the rejection of the DATA command. This is
2772 intended to be helpful for MUAs that show only the final error to their
2775 PH/42 Another patch from the Sieve maintainer.
2777 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2778 Eximstats - Fixed some broken HTML links and added missing column headers
2780 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2781 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2783 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2785 PH/43 Yet another patch from the Sieve maintainer.
2787 PH/44 I found a way to check for a TCP/IP connection going away before sending
2788 the response to the final '.' that terminates a message, but only in the
2789 case where the client has not sent further data following the '.'
2790 (unfortunately, this is allowed). However, in many cases there won't be
2791 any further data because there won't be any more messages to send. A call
2792 to select() can be used: if it shows that the input is "ready", there is
2793 either input waiting, or the socket has been closed. An attempt to read
2794 the next input character can distinguish the two cases. Previously, Exim
2795 would have sent an OK response which the client would never have see.
2796 This could lead to message repetition. This fix should cure that, at
2797 least in a lot of common cases.
2799 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2800 advertised in response to EHLO.
2806 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2807 fixed by 4.65/MH/01 (is this a record?) are fixed:
2809 (i) An empty string was always treated as zero by the numeric comparison
2810 operators. This behaviour has been restored.
2812 (ii) It is documented that the numeric comparison operators always treat
2813 their arguments as decimal numbers. This was broken in that numbers
2814 starting with 0 were being interpreted as octal.
2816 While fixing these problems I realized that there was another issue that
2817 hadn't been noticed. Values of message_size_limit (both the global option
2818 and the transport option) were treated as octal if they started with 0.
2819 The documentation was vague. These values are now always treated as
2820 decimal, and I will make that clear in the documentation.
2826 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2827 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2830 MH/01 Don't check that the operands of numeric comparison operators are
2831 integers when their expansion is in "skipping" mode (fixes bug
2832 introduced by 4.64-PH/07).
2834 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2835 child addresses, Exim now panics and dies. Previously, because the count
2836 is held in a short int, deliveries were likely to be lost. As such a
2837 large number of recipients for a single message is ridiculous
2838 (performance will be very, very poor), I have chosen to impose a limit
2839 rather than extend the field.
2845 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2846 leftover -K file (the existence of which was triggered by #402).
2847 While we were at it, introduced process PID as part of the -K
2848 filename. This should rule out race conditions when creating
2851 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2852 processing considerably. Previous code took too long for large mails,
2853 triggering a timeout which in turn triggers #401.
2855 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2856 in the DK code in transports.c. sendfile() is not really portable,
2857 hence the _LINUX specificness.
2859 TF/01 In the add_headers option to the mail command in an Exim filter,
2860 there was a bug that Exim would claim a syntax error in any
2861 header after the first one which had an odd number of characters
2864 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2865 callout verification, Exim cached a "reject" for the entire domain. This
2866 is correct for most verifications, but it is not correct for a recipient
2867 verification with use_sender or use_postmaster set, because in that case
2868 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2869 case of MAIL FROM:<> rejection from other early rejections (e.g.
2870 rejection of HELO). When verifying a recipient using a non-null MAIL
2871 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2872 Whatever the result of the callout, the value of the domain cache is
2873 left unchanged (for any other kind of callout, getting as far as trying
2874 RCPT means that the domain itself is ok).
2876 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2879 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2880 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2881 session, the daemon ignores ECONNECT errors and logs others; it now
2882 ignores EPIPE as well.
2884 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2885 (quoted-printable decoding).
2887 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2888 later the small subsequent patch to fix an introduced bug.
2890 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2892 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2894 PH/08 An error is now given if message_size_limit is specified negative.
2896 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2897 to be given (somewhat) arbitrary names.
2899 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2902 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2903 miscellaneous code fixes
2905 PH/10 Added the log_reject_target ACL modifier to specify where to log
2908 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2909 hostname. This is wrong, because it relates to the incoming message (and
2910 probably the interface on which it is arriving) and not to the outgoing
2911 callout (which could be using a different interface). This has been
2912 changed to use the value of the helo_data option from the smtp transport
2913 instead - this is what is used when a message is actually being sent. If
2914 there is no remote transport (possible with a router that sets up host
2915 addresses), $smtp_active_hostname is used.
2917 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2918 tweaks were necessary in order to get it to work (see also 21 below):
2919 (a) The code assumed that strncpy() returns a negative number on buffer
2920 overflow, which isn't the case. Replaced with Exim's string_format()
2922 (b) There were several signed/unsigned issues. I just did the minimum
2923 hacking in of casts. There is scope for a larger refactoring.
2924 (c) The code used strcasecmp() which is not a standard C function.
2925 Replaced with Exim's strcmpic() function.
2926 (d) The code set only $1; it now sets $auth1 as well.
2927 (e) A simple test gave the error "authentication client didn't specify
2928 service in request". It would seem that Dovecot has changed its
2929 interface. Fortunately there's a specification; I followed it and
2930 changed what the client sends and it appears to be working now.
2932 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2935 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2936 address A is aliased to B and C, where B exists and C does not. Without
2937 -v the output is "A verified" because verification stops after a
2938 successful redirection if more than one address is generated. However,
2939 with -v the child addresses are also verified. Exim was outputting "A
2940 failed to verify" and then showing the successful verification for C,
2941 with its parentage. It now outputs "B failed to verify", showing B's
2942 parentage before showing the successful verification of C.
2944 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2945 look up a TXT record in a specific list after matching in a combined
2948 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2949 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2950 they consult the DNS. I had assumed they would set it the way they
2951 wanted; and indeed my experiments on Linux seem to show that in some
2952 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2953 To be on the safe side, however, I have now made the interface to
2954 host_find_byname() similar to host_find_bydns(), with an argument
2955 containing the DNS resolver options. The host_find_byname() function now
2956 sets these options at its start, just as host_find_bydns() does. The smtp
2957 transport options dns_qualify_single and dns_search_parents are passed to
2958 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2959 of host_find_byname() use the default settings of RES_DEFNAMES
2960 (qualify_single) but not RES_DNSRCH (search_parents).
2962 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2963 spool_read_header() do less string testing, by means of a preliminary
2964 switch on the second character of optional "-foo" lines. (This is
2965 overdue, caused by the large number of possibilities that now exist.
2966 Originally there were few.) While I was there, I also converted the
2967 str(n)cmp tests so they don't re-test the leading "-" and the first
2968 character, in the hope this might squeeze out yet more improvement.
2970 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2971 flag allowing group syntax was set by the header_syntax check but not
2972 turned off, possible causing trouble later; (2) The flag was not being
2973 set at all for the header_verify test, causing "group"-style headers to
2974 be rejected. I have now set it in this case, and also caused header_
2975 verify to ignore an empty address taken from a group. While doing this, I
2976 came across some other cases where the code for allowing group syntax
2977 while scanning a header line wasn't quite right (mostly, not resetting
2978 the flag correctly in the right place). These bugs could have caused
2979 trouble for malformed header lines. I hope it is now all correct.
2981 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2982 with the "reply" argument non-NULL. The code, however (which originally
2983 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2984 but it didn't always do it. This confused somebody who was copying the
2985 code for some other use. I have removed all the tests.
2987 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2988 feature that was used to support insecure browsers during the U.S. crypto
2989 embargo. It requires special client support, and Exim is probably the
2990 only MTA that supported it -- and would never use it because real RSA is
2991 always available. This code has been removed, because it had the bad
2992 effect of slowing Exim down by computing (never used) parameters for the
2993 RSA_EXPORT functionality.
2995 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2996 authenticator to fail if there's a tab character in the incoming data
2997 (there should never be unless someone is messing about, as it's supposed
2998 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2999 if the connection is using TLS or if the remote IP is the same as the
3000 local IP, and the "valid-client-cert option" if a client certificate has
3003 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3004 authenticators. This can be used for authorization after authentication
3005 succeeds. (In the case of plaintext, it servers for both authentication
3008 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3009 if any retry times were supplied.
3011 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3012 connection, where there is no client IP address to check. In this
3013 situation, the verify now always succeeds.
3015 PH/25 Applied John Jetmore's -Mset patch.
3017 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3019 PH/27 In a string expansion for a processed (not raw) header when multiple
3020 headers of the same name were present, leading whitespace was being
3021 removed from all of them, but trailing whitespace was being removed only
3022 from the last one. Now trailing whitespace is removed from each header
3023 before concatenation. Completely empty headers in a concatenation (as
3024 before) are ignored.
3026 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3027 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3029 PH/29 [Removed. This was a change that I later backed out, and forgot to
3030 correct the ChangeLog entry (that I had efficiently created) before
3031 committing the later change.]
3033 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3034 address errors (4xx response to RCPT) over the same connection as other
3035 messages routed to the same hosts. Such deliveries are always "forced",
3036 so retry times are not inspected. This resulted in far too many retries
3037 for the affected addresses. The effect occurred only when there were more
3038 hosts than the hosts_max_try setting in the smtp transport when it had
3039 the 4xx errors. Those hosts that it had tried were not added to the list
3040 of hosts for which the message was waiting, so if all were tried, there
3041 was no problem. Two fixes have been applied:
3043 (i) If there are any address or message errors in an SMTP delivery, none
3044 of the hosts (tried or untried) are now added to the list of hosts
3045 for which the message is waiting, so the message should not be a
3046 candidate for sending over the same connection that was used for a
3047 successful delivery of some other message. This seems entirely
3048 reasonable: after all the message is NOT "waiting for some host".
3049 This is so "obvious" that I'm not sure why it wasn't done
3050 previously. Hope I haven't missed anything, but it can't do any
3051 harm, as the worst effect is to miss an optimization.
3053 (ii) If, despite (i), such a delivery is accidentally attempted, the
3054 routing retry time is respected, so at least it doesn't keep
3055 hammering the server.
3057 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3058 in ${readsocket because some servers need this prod.
3060 PH/32 Added some extra debug output when updating a wait-xxx database.
3062 PH/33 The hint "could be header name not terminated by colon", which has been
3063 given for certain expansion errors for a long time, was not being given
3064 for the ${if def:h_colon_omitted{... case.
3066 PH/34 The spec says: "With one important exception, whenever a domain list is
3067 being scanned, $domain contains the subject domain." There was at least
3068 one case where this was not true.
3070 PH/35 The error "getsockname() failed: connection reset by peer" was being
3071 written to the panic log as well as the main log, but it isn't really
3072 panic-worthy as it just means the connection died rather early on. I have
3073 removed the panic log writing for the ECONNRESET error when getsockname()
3076 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3077 runs only) independently of the message's sender address. This meant
3078 that, if the 4xx error was in fact related to the sender, a different
3079 message to the same recipient with a different sender could confuse
3080 things. In particular, this can happen when sending to a greylisting
3081 server, but other circumstances could also provoke similar problems.
3082 I have changed the default so that the retry time for these errors is now
3083 based a combination of the sender and recipient addresses. This change
3084 can be overridden by setting address_retry_include_sender=false in the
3087 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3088 remote server are returned as part of bounce messages. This was not
3089 happening for LMTP over a pipe (the lmtp transport), but now it is the
3090 same for both kinds of LMTP.
3092 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3093 in header lines that were in fact CNAMEs. This is no longer the case.
3095 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3096 and queue runs started by the daemon processed all messages. This has
3097 been fixed so that -R and -S can now usefully be given with -q<time>.
3099 PH/40 Import PCRE release 6.7 (fixes some bugs).
3101 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3103 PH/42 Give an error if -q is specified more than once.
3105 PH/43 Renamed the variables $interface_address and $interface_port as
3106 $received_ip_address and $received_port, to make it clear that these
3107 values apply to message reception, and not to the outgoing interface when
3108 a message is delivered. (The old names remain recognized, of course.)
3110 PH/44 There was no timeout on the connect() call when using a Unix domain
3111 socket in the ${readsocket expansion. There now is.
3113 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3114 be meaningful with "accept".
3116 SC/01 Eximstats V1.43
3117 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3119 SC/02 Eximstats V1.44
3120 Use a glob alias rather than an array ref in the generated
3121 parser. This improves both readability and performance.
3123 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3124 Collect SpamAssassin and rejection statistics.
3125 Don't display local sender or destination tables unless
3126 there is data to show.
3127 Added average volumes into the top table text output.
3129 SC/04 Eximstats V1.46
3130 Collect data on the number of addresses (recipients)
3131 as well as the number of messages.
3133 SC/05 Eximstats V1.47
3134 Added 'Message too big' to the list of mail rejection
3135 reasons (thanks to Marco Gaiarin).
3137 SC/06 Eximstats V1.48
3138 Mainlog lines which have GMT offsets and are too short to
3139 have a flag are now skipped.
3141 SC/07 Eximstats V1.49 (Alain Williams)
3142 Added the -emptyok flag.
3144 SC/08 Eximstats V1.50
3145 Fixes for obtaining the IP address from reject messages.
3147 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3148 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3149 whitespace changes from 4.64-PH/27
3151 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3154 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3155 are found, allow negative numbers in numeric criteria)
3157 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3159 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3160 to match changes made in 4.64-PH/43
3162 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3164 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3165 "domain" is actually an IP address, and force a failure. This locks out
3166 those revolvers/nameservers that support "A-for-A" lookups, in
3167 contravention of the specifications.
3169 PH/48 When a host name was looked up from an IP address, and the subsequent
3170 forward lookup of the name timed out, the host name was left in
3171 $sender_host_name, contrary to the specification.
3173 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3174 restricted to single-key lookups, Exim was not diagnosing an error if
3175 * or *@ was used with a query-style lookup.
3177 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3179 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3180 long ago, but noone interested enough thought of it. Let's just say that
3181 the "1.1" means that there are some new functions that weren't there at
3182 some point in the past.
3184 PH/51 Error processing for expansion failure of helo_data from an smtp
3185 transport during callout processing was broken.
3187 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3188 tested/used via the -bh/-bhc/-bs options.
3190 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3191 bug, fixed in subsequent PCRE releases).
3193 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3194 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3196 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3202 SC/01 Use a glob alias rather than an array ref in eximstats generated
3203 parser. This improves both readability and performance.
3205 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3206 Don't display local sender or destination tables in eximstats unless
3207 there is data to show.
3208 Added average volumes into the eximstats top table text output.
3210 SC/03 Collect data on the number of addresses (recipients) as well
3211 as the number of messages in eximstats.
3213 TF/01 Correct an error in the documentation for the redirect router. Exim
3214 does (usually) call initgroups() when daemonizing.
3216 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3217 with consistent privilege compared to when running as a daemon.
3219 TF/03 Note in the spec that $authenticated_id is not set for local
3220 submissions from trusted users.
3222 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3223 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3225 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3226 by adding some example configuration directives to the default
3227 configuration file. A little bit of work is required to uncomment the
3228 directives and define how usernames and passwords are checked, but
3229 there is now a framework to start from.
3231 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3232 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3233 without this. I don't know how relevant this is to other LDAP libraries.
3235 PH/02 Add the verb name to the "unknown ACL verb" error.
3237 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3239 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3241 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3242 directory not expanded when it should be if an expanded home directory
3243 was set for the address (which is overridden by the transport).
3245 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3248 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3249 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3250 because it is too late at that time, and has no effect.
3252 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3253 security issue with \' (bugzilla #107). I could not use the
3254 PQescapeStringConn() function, because it needs a PGconn value as one of
3257 PH/08 When testing addresses using -bt, indicate those final addresses that
3258 are duplicates that would not cause an additional delivery. At least one
3259 person was confused, thinking that -bt output corresponded to deliveries.
3260 (Suppressing duplicates isn't a good idea as you lose the information
3261 about possibly different redirections that led to the duplicates.)
3263 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3264 systems where poll() doesn't work, in particular OS X.
3266 PH/10 Added more information to debugging output for retry time not reached.
3268 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3269 operations in malware.c.
3271 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3274 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3275 syslog_duplication set false, log lines that would normally be written
3276 both the the main log and to the reject log were not written to syslog at
3279 PH/14 In the default configuration, change the use of "message" in ACL warn
3280 statements to "add_header".
3282 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3283 not followed by a command (e.g. "seen endif").
3285 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3286 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3289 PH/17 Added extra conditions to the default value of delay_warning_condition
3293 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3294 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3295 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3298 The Auto-Submitted: and various List- headers are standardised, whereas I
3299 don't think Precedence: ever was.
3301 PH/18 Refactored debugging code in route_finduser() to show more information,
3302 in particular, the error code if getpwnam() issues one.
3304 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3305 This is apparently needed in addition to the PH/07 change above to avoid
3306 any possible encoding problems.
3308 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3309 but not after initializing Perl.
3311 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3312 output them only if debugging. By default they are written stderr,
3313 apparently, which is not desirable.
3315 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3318 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3321 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3323 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3324 authenticated or an ident call has been made. Suppress the default
3325 values for $authenticated_id and $authenticated_sender (but permit -oMai
3326 and -oMas) when testing with -bh.
3328 PH/24 Re-jigged the order of the tests in the default configuration so that the
3329 tests for valid domains and recipients precede the DNS black list and CSA
3330 tests, on the grounds that those ones are more expensive.
3332 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3333 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3334 This bug exists in every version of Exim that I still have, right back to
3337 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3338 However, an attempt to turn on case-sensitivity in a regex key by
3339 including (?-i) didn't work because the subject string was already
3340 lowercased, and the effects were non-intuitive. It turns out that a
3341 one-line patch can be used to allow (?-i) to work as expected.
3347 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3348 other effects) broke the use of negated acl sub-conditions.
3350 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3353 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3354 "Deny" causes Exim to reject the incoming connection with a 554 error.
3355 Unfortunately, if there is a major crisis, such as a disk failure,
3356 tcp-wrappers gives "deny", whereas what one would like would be some
3357 kind of temporary error. A kludge has been added to help with this.
3358 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3359 554 error is used if errno is still zero or contains ENOENT (which occurs
3360 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3363 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3365 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3366 errors. Otherwise a message that provokes a temporary error (when other
3367 messages do not) can cause a whole host to time out.
3369 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3370 addresses were routed directly to files or pipes from a redirect router.
3371 File deliveries just didn't batch; pipe deliveries might have suffered
3374 PH/06 A failure to get a lock for a hints database would erroneously always say
3375 "Failed to get write lock", even when it was really a read lock.
3377 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3378 of 0600. This has been changed to use the value of the lockfile_mode
3379 option (which defaults to 0600).
3381 PH/08 Applied small patch from the Sieve maintainer.
3383 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3384 folder from quota calculations, a direct delivery into this folder messed
3385 up the contents of the maildirsize file. This was because the regex was
3386 used only to exclude .Trash (or whatever) when the size of the mailbox
3387 was calculated. There was no check that a delivery was happening into an
3388 excluded directory. This bug has been fixed by ignoring all quota
3389 processing for deliveries into excluded directories.
3391 PH/10 Added the maildirfolder_create_regex option to appendfile.
3397 PH/01 The code for finding all the local interface addresses on a FreeBSD
3398 system running IPv6 was broken. This may well have applied to all BSD
3399 systems, as well as to others that have similar system calls. The broken
3400 code found IPv4 interfaces correctly, but gave incorrect values for the
3401 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3402 that it would not match correctly against @[] and not recognize the IPv6
3405 PH/02 The ipliteral router was not recognizing addresses of the form user@
3406 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3408 PH/03 Added disable_ipv6.
3410 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3411 decoded form, because it is most often used to construct To: headers
3412 lines in autoreplies, and the decoded form may well be syntactically
3413 invalid. However, $reply_address has leading white space removed, and all
3414 newlines turned into spaces so that the autoreply transport does not
3417 PH/05 If group was specified without a user on a router, and no group or user
3418 was specified on a transport, the group from the router was ignored.
3420 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3421 for visible compile-time settings that can be used to change these
3422 numbers, for those that want even more. Backwards compatibility with old
3423 spool files has been maintained. However, going back to a previous Exim
3424 release will lost any variables that are in spool files.
3426 PH/07 Two small changes when running in the test harness: increase delay when
3427 passing a TCP/IP connection to a new process, in case the original
3428 process has to generate a bounce, and remove special handling of
3429 127.0.0.2 (sic), which is no longer necessary.
3431 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3432 be the same on different OS.
3434 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3437 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3438 whether --show-vars was specified or not
3440 JJ/02 exipick: Added support for new ACL variable spool format introduced
3443 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3444 syntactically invalid From: or Reply-to: line, and a filter used this to
3445 generate an autoreply, and therefore failed to obtain an address for the
3446 autoreply, Exim could try to deliver to a non-existent relative file
3447 name, causing unrelated and misleading errors. What now happens is that
3448 it logs this as a hard delivery error, but does not attempt to create a
3451 PH/11 The exinext utility has a -C option for testing purposes, but although
3452 the given file was scanned by exinext itself; it wasn't being passed on
3453 when Exim was called.
3455 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3456 an end-of-file indication when reading a command response.
3458 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3459 compiled. In many other places in Exim, IPv6 addresses are always
3460 recognized, so I have changed this. It also means that IPv4 domain
3461 literals of the form [IPV4:n.n.n.n] are now always recognized.
3463 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3464 used if the router is not running as root, for example, when verifying at
3465 ACL time, or when using -bh. The debugging output from this situation was
3466 non-existent - all you got was a failure to exec. I have made two
3469 (a) Failures to set uid/gid, the current directory, or a process leader
3470 in a subprocess such as that created by queryprogram now generate
3471 suitable debugging output when -d is set.
3473 (b) The queryprogram router detects when it is not running as root,
3474 outputs suitable debugging information if -d is set, and then runs
3475 the subprocess without attempting to change uid/gid.
3477 PH/15 Minor change to Makefile for building test_host (undocumented testing
3480 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3481 additional section of a DNS packet that returns MX or SRV records.
3482 Instead, it always explicitly searches for A/AAAA records. This avoids
3483 major problems that occur when a DNS server includes only records of one
3484 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3485 fixed another bug: if SRV records were looked up and the corresponding
3486 address records were *not* found in the additional section, the port
3487 values from the SRV records were lost.
3489 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3490 using the correct key (the original address) when searching the retry
3491 rules in order to find which one to use for generating the retry hint.
3493 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3494 adding the default one. Similarly, if it contains a Reply-To: header, the
3495 errors_reply_to option, if set, is not used.
3497 PH/19 When calculating a retry time, Exim used to measure the "time since
3498 failure" by looking at the "first failed" field in the retry record. Now
3499 it does not use this if it is later than than the arrival time of the
3500 message. Instead it uses the arrival time. This makes for better
3501 behaviour in cases where some deliveries succeed, thus re-setting the
3502 "first failed" field. An example is a quota failure for a huge message
3503 when small messages continue to be delivered. Without this change, the
3504 "time since failure" will always be short, possible causing more frequent
3505 delivery attempts for the huge message than are intended.
3506 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3508 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3509 $1, $2, $3) because the numerical variables can be reset during some
3510 expansion items (e.g. "match"), thereby losing the authentication data.
3512 PH/21 Make -bV show the size of off_t variables so that the test suite can
3513 decide whether to run tests for quotas > 2G.
3515 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3516 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3517 filecount value is greater than 2G or if a quota value is greater than 2G
3518 on a system where the size of off_t is not greater than 4, a panic error
3521 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3522 never match. The debug and -bh output now contains an explicit error
3523 message indicating a malformed IPv4 address or mask.
3525 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3526 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3527 PH/23 above applies.
3529 PH/25 Do not write to syslog when running in the test harness. The only
3530 occasion when this arises is a failure to open the main or panic logs
3531 (for which there is an explicit test).
3533 PH/26 Added the /no_tell option to "control=freeze".
3535 PH/27 If a host name lookup failed very early in a connection, for example, if
3536 the IP address matched host_lookup and the reverse lookup yielded a name
3537 that did not have a forward lookup, an error message of the form "no IP
3538 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3539 could be logged. Now it outputs the IP address instead of "NULL".
3541 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3542 allows the sender and the authenticated sender to be set when
3543 submitting a message from within Exim. Since child_open_exim() is
3544 documented for local_scan(), the new function should be too.
3546 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3547 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3548 results in an empty string is now treated as unset.
3550 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3552 PH/31 Added sender_verify_fail logging option.
3554 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3555 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3556 tidied the source and removed it altogether.
3558 PH/33 When a queue run was abandoned because the load average was too high, a
3559 log line was always written; now it is written only if the queue_run log
3560 selector is set. In addition, the log line for abandonment now contains
3561 information about the queue run such as the pid. This is always present
3562 in "start" and "stop" lines but was omitted from the "abandon" line.
3564 PH/34 Omit spaces between a header name and the colon in the error message that
3565 is given when verify = headers_syntax fails (if there are lots of them,
3566 the message gets confusing).
3568 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3569 names, as there are now some PTR records that contain slashes. This check
3570 is only to protect against broken name servers that fall over on strange
3571 characters, so the fact that it applies to all lookups doesn't matter.
3573 PH/36 Now that the new test suite is complete, we can remove some of the
3574 special code in Exim that was needed for the old test suite. For example,
3575 sorting DNS records because real resolvers return them in an arbitrary
3576 order. The new test suite's fake resolver always returns records in the
3579 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3580 bounces) except when queue_only is set, to avoid logging races between
3581 the different processes.
3583 PH/38 Panic-die if .include specifies a non-absolute path.
3585 PH/39 A tweak to the "H" retry rule from its user.
3587 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3588 a label. They prevented compilation on older perls.
3590 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3591 a warning to be raised on newish perls.
3593 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3594 on queue. Changes to match documented behaviour of showing count of
3595 messages matching specified criteria.
3597 PH/40 Changed the default ident timeout from 30s to 5s.
3599 PH/41 Added support for the use of login_cap features, on those BSD systems
3600 that have them, for controlling the resources used by pipe deliveries.
3602 PH/42 The content-scanning code uses fopen() to create files in which to put
3603 message data. Previously it was not paying any attention to the mode of
3604 the files. Exim runs with umask(0) because the rest of the code creates
3605 files with open(), and sets the required mode explicitly. Thus, these
3606 files were ending up world-writeable. This was not a big issue, because,
3607 being within the spool directory, they were not world-accessible. I have
3608 created a function called modefopen, which takes an additional mode
3609 argument. It sets umask(777), creates the file, chmods it to the required
3610 mode, then resets the umask. All the relevant calls to fopen() in the
3611 content scanning code have been changed to use this function.
3613 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3614 to 24 hours. This avoids potential overflow problems when processing G
3615 and H retry rules. I suspect nobody ever tinkers with this value.
3617 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3619 PH/45 When the plaintext authenticator is running as a client, the server's
3620 challenges are checked to ensure they are valid base64 strings. By
3621 default, the authentication attempt is cancelled if an invalid string is
3622 received. Setting client_ignore_invalid_base64 true ignores these errors.
3623 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3624 they are received. Thus, the responses can be made to depend on the
3625 challenges. If an invalid string is ignored, an empty string is placed in
3628 PH/46 Messages that are created by the autoreply transport now contains a
3629 References: header, in accordance with RFCs 2822 and 3834.
3631 PH/47 Added authenticated_sender_force to the smtp transport.
3633 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3635 PH/49 Installed latest patch from the Sieve maintainer.
3637 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3638 was also set, the appendfile transport was unnecessarily scanning a
3639 directory of message files (e.g. for maildir delivery) to find the count
3640 of files (along with the size), even though it did not need this
3641 information. It now does the scan only if it needs to find either the
3642 size of the count of files.
3644 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3646 PH/52 Two bugs concerned with error handling when the smtp transport is
3649 (i) Exim was not creating retry information for temporary errors given
3650 for individual recipients after the DATA command when the smtp transport
3651 was used in LMTP mode. This meant that they could be retried too
3652 frequently, and not timed out correctly.
3654 (ii) Exim was setting the flag that allows error details to be returned
3655 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3656 recipients that were returned after the DATA command.
3658 PH/53 This is related to PH/52, but is more general: for any failing address,
3659 when detailed error information was permitted to be returned to the
3660 sender, but the error was temporary, then after the final timeout, only
3661 "retry timeout exceeded" was returned. Now it returns the full error as
3662 well as "retry timeout exceeded".
3664 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3665 do this, and (what is worse) MTAs that accept it.
3667 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3668 will now be deprecated.
3670 PH/56 New os.c-cygwin from the Cygwin maintainer.
3672 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3673 formats (previously only available in exim formats via -bpr, -bpru,
3674 and -bpra. Now also available in native and exiqgrep formats)
3676 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3677 with very large, slow to parse queues
3679 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3681 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3683 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3684 responses to authentication challenges, though it was showing the
3685 challenges; (ii) I've removed the CR characters from the debug output for
3688 PH/58 Allow for the insertion of a newline as well as a space when a string
3689 is turned into more than one encoded-word during RFC 2047 encoding. The
3690 Sieve code now uses this.
3692 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3693 data_4xx, lost_connection, tls_required.
3695 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3696 message was being sent as an SMTP response.
3698 PH/61 Add -l and -k options to exicyclog.
3700 PH/62 When verifying, if an address was redirected to one new address, so that
3701 verification continued, and the new address failed or deferred after
3702 having set something in $address_data, the value of $address_data was not
3703 passed back to the ACL. This was different to the case when no
3704 redirection occurred. The value is now passed back in both cases.
3706 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3707 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3708 use login_cap.h, so on its own it isn't the distinguishing feature. The
3709 new name refers directly to the setclassresources() function.
3711 PH/65 Added configuration files for NetBSD3.
3713 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3715 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3716 is preferred over IPv4.
3718 PH/68 The bounce_return_message and bounce_return_body options were not being
3719 honoured for bounces generated during the reception of non-SMTP messages.
3720 In particular, this applied to messages rejected by the ACL. This bug has
3721 been fixed. However, if bounce_return_message is true and bounce_return_
3722 body is false, the headers that are returned for a non-SMTP message
3723 include only those that have been read before the error was detected.
3724 (In the case of an ACL rejection, they have all been read.)
3726 PH/69 The HTML version of the specification is now built in a directory called
3727 spec_html instead of spec.html, because the latter looks like a path with
3728 a MIME-type, and this confuses some software.
3730 PH/70 Catch two compiler warnings in sieve.c.
3732 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3733 function verify_get_ident() calls ip_connect() to connect a socket, but
3734 if the "connect()" function timed out, ip_connect() used to close the
3735 socket. However, verify_get_ident() also closes the socket later, and in
3736 between Exim writes to the log, which may get opened at this point. When
3737 the socket was closed in ip_connect(), the log could get the same file
3738 descriptor number as the socket. This naturally causes chaos. The fix is
3739 not to close the socket in ip_connect(); the socket should be closed by
3740 the function that creates it. There was only one place in the code where
3741 this was missing, in the iplookup router, which I don't think anybody now
3742 uses, but I've fixed it anyway.
3744 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3745 well as to direct DNS lookups. Otherwise the handling of names in host
3746 lists is inconsistent and therefore confusing.
3752 PH/01 Two changes to the default runtime configuration:
3754 (1) Move the checks for relay_from_hosts and authenticated clients from
3755 after to before the (commented out) DNS black list checks.
3757 (2) Add control=submission to the relay_from_hosts and authenticated
3758 clients checks, on the grounds that messages accepted by these
3759 statements are most likely to be submissions.
3761 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3763 (1) Generate an error if the third argument for the ${prvs expansion is
3766 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3769 (3) Reset the variables that are obtained from the first argument of
3770 ${prvscheck and used in the second argument before leaving the code,
3771 because their memory is reclaimed, so using them afterwards may do
3774 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3775 one (it's much easier than Tom thought :-).
3777 (5) Because of (4), we can now allow for the use of $prvscheck_result
3778 inside the third argument.
3780 PH/03 For some reason, the default setting of PATH when running a command from
3781 a pipe transport was just "/usr/bin". I have changed it to
3784 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3785 anything to be listed in the output from -bV.
3787 PH/05 When a filter generated an autoreply, the entire To: header line was
3788 quoted in the delivery log line, like this:
3790 => >A.N.Other <ano@some.domain> <original@ddress> ...
3792 This has been changed so that it extracts the operative address. There
3793 may be more than one such address. If so, they are comma-separated, like
3796 => >ano@some.domain,ona@other.domain <original@ddress> ...
3798 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3799 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3800 not being looked up in the rDNS to get a host name, Exim was showing the
3801 IP address twice in Received: lines, even though the IP addresses were
3802 identical. For example:
3804 Received: from [1.2.3.4] (helo=[1.2.3.4])
3806 However, if the real host name was known, it was omitting the HELO data
3807 if it matched the actual IP address. This has been tidied up so that it
3808 doesn't show the same IP address twice.
3810 PH/07 When both +timestamp and +memory debugging was on, the value given by
3811 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3812 called by the debug printing, thereby overwriting the timestamp buffer.
3813 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3815 PH/08 When the original message was included in an autoreply transport, it
3816 always said "this is a copy of the message, including all the headers",
3817 even if body_only or headers_only was set. It now gives an appropriate
3820 PH/09 Applied a patch from the Sieve maintainer which:
3822 o fixes some comments
3823 o adds the (disabled) notify extension core
3824 o adds some debug output for the result of if/elsif tests
3825 o points to the current vacation draft in the documentation
3826 and documents the missing references header update
3830 o fixes a bug in processing the envelope test (when testing
3831 multiple envelope elements, the last element determined the
3834 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3835 Electronic Mail") by including:
3837 Auto-submitted: auto-generated
3839 in the messages that it generates (bounce messages and others, such as
3840 warnings). In the case of bounce messages for non-SMTP messages, there was
3841 also a typo: it was using "Auto_submitted" (underscore instead of
3842 hyphen). Since every message generated by Exim is necessarily in response
3843 to another message, thes have all been changed to:
3845 Auto-Submitted: auto-replied
3847 in accordance with these statements in the RFC:
3849 The auto-replied keyword:
3851 - SHOULD be used on messages sent in direct response to another
3852 message by an automatic process,
3854 - MUST NOT be used on manually-generated messages,
3856 - MAY be used on Delivery Status Notifications (DSNs) and Message
3857 Disposition Notifications (MDNs),
3859 - MUST NOT be used on messages generated by automatic or periodic
3860 processes, except for messages which are automatic responses to
3863 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3864 to the default Received: header definition.
3866 PH/12 Added log selector acl_warn_skipped (default on).
3868 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3869 variables because (a) they are in the wrong storage pool and (b) even if
3870 they were copied, it wouldn't work properly because of the caching.
3872 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3873 checking when decoding. Apparently there are clients that generate
3874 overlong encoded strings. Why am I not surprised?
3876 PH/15 If the first argument of "${if match_address" was not empty, but did not
3877 contain an "@" character, Exim crashed. Now it writes a panic log message
3878 and treats the condition as false.
3880 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3882 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3883 extension "envelope-auth". The code is finished and in agreement with
3884 other implementations, but there is no documentation so far and in fact,
3885 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3886 not changing the active code.
3888 Print executed "if" and "elsif" statements when debugging is used. This
3889 helps a great deal to understand what a filter does.
3891 Document more things not specified clearly in RFC3028. I had all this
3892 sorted out, when out of a sudden new issues came to my mind. Oops."
3894 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3897 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3898 canonical form (as documented). However, after a host name lookup from
3899 the IP address, check_host() was doing a simple string comparison with
3900 addresses acquired from the DNS when checking that the found name did
3901 have the original IP as one of its addresses. Since any found IPv6
3902 addresses are likely to be in abbreviated form, the comparison could
3903 fail. Luckily, there already exists a function for doing the comparison
3904 by converting both addresses to binary, so now that is used instead of
3905 the text comparison.
3907 PH/20 There was another similar case to PH/19, when a complete host name was
3908 given in a host list; looking up its IP address could give an abbreviated
3909 form, whereas the current host's name might or might not be abbreviated.
3910 The same fix has been applied.
3916 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3917 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3920 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3921 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3923 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3925 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3926 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3927 Also, for "Auto-Submitted", treat anything other than "no" as
3928 non-personal, in accordance with RFC 3834. (Previously it treated
3929 anything starting "auto-" as non-personal.)
3931 TF/01 The control=submission/name=... option had a problem with syntax
3932 errors if the name included a slash character. The /name= option
3933 now slurps the rest of the string, so it can include any characters
3934 but it must come last in the list of options (after /sender_retain
3937 PH/05 Some modifications to the interface to the fake nameserver for the new
3945 TK/01 Added the "success_on_redirect" address verification option. See
3946 NewStuff for rationale and an example.
3948 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3950 PH/02 Patch to exigrep to allow it to work on syslog lines.
3952 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3953 fread() to skip over the body file's header line, because in Cygwin the
3954 header line is locked and is inaccessible.
3956 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3957 co-exist for some time) to make it clear that it is the Exim ID that is
3958 referenced, not the Message-ID: header line.
3960 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3961 string_format() function, because snprintf() does not exist on all
3964 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3965 ${stat: expansion item.
3967 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3968 protocol synchronization error", to keep the pedants happy.
3970 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3971 well as for IRIX systems, when gcc is being used. See the host.c source
3974 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3976 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3979 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3980 transport and to the smtp transport in LMTP mode.
3982 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3984 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3985 run a filter in a subprocess. This could lead to confusion in subsequent
3986 lookups in the parent process. There should also be a search_tidyup() at
3987 the end of the subprocess.
3989 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3990 only if the host matched helo_try_verify_hosts, which caused the
3991 verification to occur when the EHLO/HELO command was issued. The ACL just
3992 tested the remembered result. Now, if a previous verification attempt has
3993 not happened, "verify = helo" does it there and then.
3995 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3997 TK/03 Fix log output including CR from clamd.
3999 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4000 exist provoked a memory error which could cause a segfault.
4002 PH/15 Installed PCRE 6.2
4004 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4006 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4007 of the problem. Specifically, suggested +O2 rather than +O1 for the
4010 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4012 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4013 record, the port was not correctly logged when the outgoing_port log
4014 selector was set (it logged the transort's default port).
4016 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4017 fallback_hosts, and "hosts" in the smtp transport.
4019 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4020 host errors such as "Connection refused".
4022 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4023 authentication with radiusclient 0.4.9:
4025 - Error returned from rc_read_config was caught wrongly
4026 - Username/password not passed on to radius server due to wrong length.
4028 The presumption is that some radiusclient API changes for 4.51/PH/17
4029 were not taken care of correctly. The code is still untested by me (my
4030 Linux distribution still has 0.3.2 of radiusclient), but it was
4031 contributed by a Radius user.
4033 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4034 expanding the "port" option of the smtp transport.
4036 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4037 while reading a MIME header. Thanks to Tom Hughes for a patch.
4039 PH/24 Include config.h inside local_scan.h so that configuration settings are
4042 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4043 that in an ACL for RCPT (for example), you can examine exactly what was
4046 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4047 commands, but it was not correctly comparing the address with the actual
4048 client host address. Thus, it would show the EHLO address in Received:
4049 header lines when this was not necessary.
4051 PH/27 Added the % operator to ${eval:}.
4053 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4054 it should be ignoring failures (because with -C, for example, it has lost
4055 privilege). It wasn't ignoring creation failures other than "already
4058 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4061 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4062 because an input line was too long, either on its own, or by virtue of
4063 too many continuations, the temporary file was not being removed, and the
4064 return code was incorrect.
4066 PH/31 Missing "BOOL" in function definition in filtertest.c.
4068 PH/32 Applied Sieve patches from the maintainer.
4070 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4072 PH/33 Added "verify = not_blind".
4074 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4075 Local/Makefile (with some defaults set). These are used in built scripts
4076 such as exicyclog, but they have never been used in the exim_install
4077 script (though there are many overriding facilities there). I have
4078 arranged that the exim_install script now takes note of these two
4081 PH/35 Installed configuration files for Dragonfly.
4083 PH/36 When a locally submitted message by a trusted user did not contain a
4084 From: header, and the sender address was obtained from -f or from an SMTP
4085 MAIL command, and the trusted user did not use -F to supply a sender
4086 name, $originator_name was incorrectly used when constructing a From:
4087 header. Furthermore, $originator_name was used for submission mode
4088 messages from external hosts without From: headers in a similar way,
4089 which is clearly wrong.
4091 PH/37 Added control=suppress_local_fixups.
4093 PH/38 When log_selector = +received_sender was set, and the addition of the
4094 sender made the log line's construction buffer exactly full, or one byte
4095 less than full, an overflow happened when the terminating "\n" was
4098 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4099 when the result of a list match is failure because a DNS lookup failed.
4101 PH/40 RM_COMMAND is now used in the building process.
4103 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4104 the "build-* directories that it finds.
4106 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4107 address in a domain literal was a prefix of an interface address.
4109 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4110 when verifying a sender address, unless rewrite_headers is false.
4112 PH/44 Wrote a long comment about why errors_to addresses are verified as
4113 recipients, not senders.
4115 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4116 the ratelimit ACL was added.
4118 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4120 PH/46 Added extra information about PostgreSQL errors to the error string.
4122 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4123 avoiding the need to install special zones in a real server. This is
4124 backwards compatible; if it can't find the fake resolver, it drops back.
4125 Thus, both old and new test suites can be run.
4127 TF/02 Added util/ratelimit.pl
4129 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4130 clock is set back in time.
4132 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4133 Candler <B.Candler@pobox.com>.
4135 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4136 OK for addresses that are the result of redirections.
4138 PH/48 A number of further additions for the benefit of the new test suite,
4139 including a fake gethostbyname() that interfaces to the fake DNS resolver
4142 TF/06 The fix for widen_domains has also been applied to qualify_single and
4143 search_parents which are the other dnslookup options that can cause
4146 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4149 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4151 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4152 service" if it encountered an MX record with an empty target hostname.
4153 The message is now "an MX or SRV record indicated no SMTP service".
4155 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4156 if the DNS of the sending site is misconfigured. This is quite a
4157 common situation. This change restores the behaviour of treating a
4158 helo verification defer as a failure.
4160 PH/51 If self=fail was set on a router, the bounce message did not include the
4161 actual error message.
4167 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4169 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4170 command itself ended in error, the underlying message about the transport
4171 filter timeout was being overwritten with the pipe command error. Now the
4172 underlying error message should be appended to the second error message.
4174 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4176 PH/02 Reduce the amount of output that "make" produces by default. Full output
4177 can still be requested.
4179 PH/03 The warning log line about a condition test deferring for a "warn" verb
4180 was being output only once per connection, rather than after each
4181 occurrence (because it was using the same function as for successful
4182 "warn" verbs). This seems wrong, so I have changed it.
4184 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4185 it should not have, which might have caused a crash in the right
4186 circumstances, but probably never did.
4188 PH/04 Installed a modified version of Tony Finch's patch to make submission
4189 mode fix the return path as well as the Sender: header line, and to
4190 add a /name= option so that you can make the user's friendly name appear
4193 TF/03 Added the control = fakedefer ACL modifier.
4195 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4196 Mark Lowes for thorough testing.
4198 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4200 TK/03 Merged latest SRS patch from Miles Wilton.
4202 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4203 with the definition in sysexits.h (which is #included earlier).
4204 Fortunately, Exim does not actually use EX_OK. The code used to try to
4205 preserve the sysexits.h value, by assuming that macro definitions were
4206 scanned for macro replacements. I have been disabused of this notion,
4207 so now the code just undefines EX_OK before #including unistd.h.
4209 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4210 in the smtp transport. When a block could not be written in a single
4211 write() function, the timeout was being re-applied to each part-write.
4212 This seems wrong - if the receiver was accepting one byte at a time it
4213 would take for ever. The timeout is now adjusted when this happens. It
4214 doesn't have to be particularly precise.
4216 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4217 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4219 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4220 without a domain if the check to <postmaster@domain> fails.
4222 SC/01 Eximstats: added -xls and the ability to specify output files
4223 (patch written by Frank Heydlauf).
4225 SC/02 Eximstats: use FileHandles for outputting results.
4227 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4229 SC/04 Eximstats: fixed display of large numbers with -nvr option
4231 SC/05 Eximstats: fixed merging of reports with empty tables.
4233 SC/06 Eximstats: added the -include_original_destination flag
4235 SC/07 Eximstats: removed tabs and trailing whitespace.
4237 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4239 TK/06 MBOX spool code: Add real "From " MBOX separator line
4240 so the .eml file is really in mbox format (even though
4241 most programs do not really care). Patch from Alex Miller.
4243 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4244 The latter is generated from $received_to and is only set if the
4245 message has one envelope recipient. SA can use these headers,
4246 obviously out-of-the-box. Patch from Alex Miller.
4248 PH/08 The ${def test on a variable was returning false if the variable's
4249 value was "0", contrary to what the specification has always said!
4250 The result should be true unless the variable is empty.
4252 PH/09 The syntax error of a character other than { following "${if
4253 def:variable_name" (after optional whitespace) was not being diagnosed.
4254 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4255 accidental colon was present, for example, could give incorrect results.
4257 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4258 result is used (not including appendfile, where other changes are about
4261 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4262 This involved changing a lot of size variables from int to off_t. It
4263 should work with maildirs and everything.
4265 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4266 spamd dying while we are connected to it.
4268 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4271 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4272 function for BDB 4.3.
4274 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4276 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4277 the sources for PCRE have been re-arranged and more files are now
4280 PH/15 The code I had for printing potentially long long variables in PH/11
4281 above was not the best (it lost precision). The length of off_t variables
4282 is now inspected at build time, and an appropriate printing format (%ld
4283 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4284 to be "long long int" or "long int". This is needed for the internal
4285 formatting function string_vformat().
4287 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4288 the configuration file to be ":syslog", then the script "guesses" where
4289 the logs files are, rather than using the compiled in default. In our
4290 case the guess is not the same as the compiled default, so the script
4291 suddenly stopped working when I started to use syslog. The patch checks
4292 to see if log_file_path is "". If so, it attempts to read it from exim
4293 with no configuration file to get the compiled in version, before it
4294 falls back to the previous guessing code."
4296 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4297 implementing BATV in an Exim configuration. See NewStuff for the gory
4300 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4301 Makefile that are specific to HP-UX.
4303 PH/18 If the "use_postmaster" option was set for a recipient callout together
4304 with the "random" option, the postmaster address was used as the MAIL
4305 FROM address for the random test, but not for the subsequent recipient
4306 test. It is now used for both.
4308 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4309 patch removes a few documentation additions to RFC 3028, because the
4310 latest draft now contains them. It adds the new en;ascii-case comparator
4311 and a new error check for 8bit text in MIME parts. Comparator and
4312 require names are now matched exactly. I enabled the subaddress
4313 extension, but it is not well tested yet (read: it works for me)."
4315 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4316 rework some of the code of TK/09 above to avoid the hardwired use of
4317 "%lld" and "long long". Replaced the call to snprintf() with a call to
4320 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4321 records point to non-existent hosts", "retry timeout exceeded", and
4322 "retry time not reached for any host after a long failure period".
4324 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4325 experimental DomainKeys support:
4327 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4328 (2) On an error such as an illegally used "control", the wrong name for
4329 the control was given.
4331 These problems did NOT occur unless DomainKeys support was compiled.
4333 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4335 PH/24 Added ${if match_ip condition.
4337 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4338 databases so that it will be absolutely obvious if a crash occurs in the
4339 DB library. This is a regular occurrence (often caused by mis-matched
4342 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4343 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4344 picked up on a user's system that detects such things. There doesn't seem
4345 to be a gcc warning option for this - only an attribute that has to be
4346 put on the function's prototype. It seems that in Fedora Core 4 they have
4347 set this on a number of new functions. No doubt there will be more in due
4350 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4351 specify a transport. However, if an address that was verified by such a
4352 router was the subject of a callout, Exim crashed because it tried to
4353 read the rcpt_include_affixes from the non-existent transport. Now it
4354 just assumes that the setting of that option is false. This bug was
4355 introduced by 4.51/PH/31.
4357 PH/28 Changed -d+all to exclude +memory, because that information is very
4358 rarely of interest, but it makes the output a lot bigger. People tend to
4359 do -d+all out of habit.
4361 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4362 code in os-type was giving problems when libc.so lives in lib64, like on
4365 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4366 aren't the modern standard, and it seems that some systems' include files
4367 don't always have them. Exim was already checking for some of the newer
4368 ones like T_AAAA, and defining it itself. I've added checks for all the
4369 record types that Exim uses.
4371 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4372 not automatically generating a new one, as it is supposed to. This
4373 prevented TLS from working. If the file did exist, but contained invalid
4374 data, a new version was generated, as expected. It was only the case of a
4375 non-existent file that was broken.
4377 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4378 with a change in libdomainkeys > 0.64.
4380 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4381 from DNS. If the selector record carries the flag, it now has
4382 precedence over the domain-wide flag.
4384 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4386 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4387 the use of a port name that isn't defined in /etc/services) occurred, the
4388 message was deferred as in a normal delivery, and thus remained on the
4389 spool, instead of being failed because of the mua_wrapper setting. This
4390 is now fixed, and I tidied up some of the mua_wrapper messages at the
4393 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4394 the messages in a hash of arrays rather than using individual hashes.
4395 This is a bit cleaner and results in dramatic memory savings, albeit
4396 at a slight CPU cost.
4398 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4399 as requested by Marc Sherman.
4401 SC/10 Eximstats: added histograms for user specified patterns as requested
4404 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4406 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4407 fopen() in the content-scanning modules that did not already have it.
4413 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4414 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4416 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4418 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4420 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4421 to test Sieve filters that use "vacation".
4423 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4424 that changes the way the GnuTLS parameters are stored in the cache file.
4425 The new format can be generated externally. For backward compatibility,
4426 if the data in the cache doesn't make sense, Exim assumes it has read an
4427 old-format file, and it generates new data and writes a new file. This
4428 means that you can't go back to an older release without removing the
4431 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4432 work if there are any delivery delays because "one_time" forces the
4433 parent to be marked "delivered", so its unseen clone is never tried
4434 again. For this reason, Exim now forbids the simultaneous setting of
4437 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4438 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4439 entry if you want to know the details. The fix, however, neglected to
4440 consider the case when local delivery batching is involved. The test for
4441 "previously delivered" was not happening when checking to see if an
4442 address could be batched with a previous (undelivered) one; under
4443 certain circumstances this could lead to multiple deliveries to the same
4446 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4447 in its include files, and this causes problems building Exim.
4449 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4450 header_syntax) but Exim was just ignoring anything given after a slash.
4451 In particular, this caused confusion with an attempt to use "verify =
4452 reverse_host_lookup/defer_ok". An error is now given when options are
4453 supplied for verify items that do not have them. (Maybe reverse_host_
4454 lookup should have a defer_ok option, but that's a different point.)
4456 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4457 defined by RFC 821) to 2048, because there were problems with some AUTH
4458 commands, and RFC 1869 says the size should be increased for extended
4459 SMTP commands that take arguments.
4461 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4464 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4465 "unknown" error; now it says that the functionality isn't in the binary.
4467 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4468 an address' error message when a string expansion fails (syntax or
4469 whatever). Otherwise the password may appear in the log. Following change
4470 PH/42 below, there is no longer a chance of it appearing in a bounce
4473 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4475 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4476 of its addresses was ever tried. (Bugzilla bug #2.)
4478 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4479 the result incorrectly in the debug output. (It correctly added a newline
4480 to what was transported.)
4482 TF/01 Added $received_time.
4484 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4485 commented out examples of how to interface to a virus scanner and to
4486 SpamAssassin. Also added commented examples of av_scanner and
4487 spamd_address settings.
4489 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4490 and controls are allowed in which ACLs. There were a couple of minor
4491 errors. Some of the entries in the conditions table (which is a table of
4492 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4493 negation of where the condition IS allowed.
4495 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4497 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4498 header file does not have a version number, so I've had to invent a new
4499 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4500 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4501 radiusclient), but it was contributed by a Radius user.
4503 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4504 files or directories, for OpenSSL.
4506 PH/19 When an Exim process that is running as root has to create an Exim log
4507 file, it does so in a subprocess that runs as exim:exim so as to get the
4508 ownership right at creation (otherwise, other Exim processes might see
4509 the file with the wrong ownership). There was no test for failure of this
4510 fork() call, which would lead to the process getting stuck as it waited
4511 for a non-existent subprocess. Forks do occasionally fail when resources
4512 run out. I reviewed all the other calls to fork(); they all seem to check
4515 PH/20 When checking for unexpected SMTP input at connect time (before writing
4516 the banner), Exim was not dealing correctly with a non-positive return
4517 from the read() function. If the client had disconnected by this time,
4518 the result was a log entry for a synchronization error with an empty
4519 string after "input=" when read() returned zero. If read() returned -1
4520 (an event I could not check), uninitialized data bytes were printed.
4521 There were reports of junk text (parts of files, etc) appearing after
4524 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4526 PH/22 Added support for macro redefinition, and (re)definition in between
4527 driver and ACL definitions.
4529 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4530 forgetting to use the resulting value; it was using the unexpanded value.
4532 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4533 hadn't been configured. The fix is from Juergen Kreileder, who
4534 understands it better than I do:
4536 "Here's what I see happening with three configured cyrus_sasl
4537 authenticators configured (plain, login, cram-md5):
4539 On startup auth_cyrus_sasl_init() gets called for each of these.
4540 This means three calls to sasl_listmech() without a specified mech_list.
4541 => SASL tests which mechs of all available mechs actually work
4542 => three warnings about OTP not working
4543 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4545 With the patch, sasl_listmech() also gets called three times. But now
4546 SASL's mech_list option is set to the server_mech specified in the the
4547 authenticator. Or in other words, the answer from sasl_listmech()
4548 gets limited to just the mech you're testing for (which is different
4550 => the return list contains just 'plain' or 'login', 'cram-md5' or
4551 nothing depending on the value of ob->server_mech.
4553 I've just tested the patch: Authentication still works fine,
4554 unavailable mechs specified in the exim configuration are still
4555 caught, and the auth.log warnings about OTP are gone."
4557 PH/25 When debugging is enabled, the contents of the command line are added
4558 to the debugging output, even when log_selector=+arguments is not
4561 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4562 answer is "GNU", and only if the return is "GNU/something" is the answer
4565 PH/27 $acl_verify_message is now set immediately after the failure of a
4566 verification in an ACL, and so is available in subsequent modifiers. In
4567 particular, the message can be preserved by coding like this:
4569 warn !verify = sender
4570 set acl_m0 = $acl_verify_message
4572 Previously, $acl_verify_message was set only while expanding "message"
4573 and "log_message" when a very denied access.
4575 PH/28 Modified OS/os.c-Linux with
4577 -#ifndef OS_LOAD_AVERAGE
4578 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4580 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4581 nomenclature these days.)
4583 PH/29 Installed patch from the Sieve maintainer that adds the options
4584 sieve_useraddress and sieve_subaddress to the redirect router.
4586 PH/30 In these circumstances:
4587 . Two addresses routed to the same list of hosts;
4588 . First host does not offer TLS;
4589 . First host accepts first address;
4590 . First host gives temporary error to second address;
4591 . Second host offers TLS and a TLS session is established;
4592 . Second host accepts second address.
4593 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4594 and peerdn, if requested) that were in fact used only for the second
4597 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4598 attention to any local part prefix or suffix that was matched by the
4599 router that accepted the address. It now behaves in the same way as it
4600 does for delivery: the affixes are removed from the local part unless
4601 rcpt_include_affixes is set on the transport.
4603 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4604 timeout during the DATA phase of an incoming message.
4606 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4607 applied a patch sanctioned by the Sieve maintainer.
4609 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4610 the uid or gid is negative. A case of a negative gid caused this to be
4611 noticed. The fix allows for either to be negative.
4613 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4614 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4617 PH/36 The change PH/12 above was broken. Fixed it.
4619 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4620 the grounds that routing the same address twice would always produce the
4621 same answer. This might have been true once, but it is certainly no
4622 longer true now. Routing a child address may depend on the previous
4623 routing that produced that child. Some complicated redirection strategies
4624 went wrong when messages had multiple recipients, and made Exim's
4625 behaviour dependent on the order in which the addresses were given.
4627 I have moved the duplicate checking until after the routing is complete.
4628 Exim scans the addresses that are assigned to local and remote
4629 transports, and removes any duplicates. This means that more work will be
4630 done, as duplicates will always all be routed, but duplicates are
4631 presumably rare, so I don't expect this is of any significance.
4633 For deliveries to pipes, files, and autoreplies, the duplicate checking
4634 still happens during the routing process, since they are not going to be
4637 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4638 It corrects a timeout issue with spamd. This is Ian's comment: "The
4639 background is that sometimes spamd either never reads data from a
4640 connection it has accepted, or it never writes response data. The exiscan
4641 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4642 blindly assumes that writes won't block so it may never time out."
4644 PH/39 Allow G after quota size as well as K and M.
4646 PH/40 The value set for $authenticated_id in an authenticator may not contain
4647 binary zeroes or newlines because the value is written to log lines and
4648 to spool files. There was no check on this. Now the value is run through
4649 the string_printing() function so that such characters are converted to
4650 printable escape sequences.
4652 PH/41 $message_linecount is a new variable that contains the total number of
4653 lines in the message. Compare $body_linecount, which is the count for the
4656 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4657 bounce and delay warning messages, except in certain special cases, which
4660 (a) An SMTP error message from a remote host;
4661 (b) A message specified in a :fail: redirection;
4662 (c) A message specified in a "fail" command in a system filter;
4663 (d) A message specified in a FAIL return from the queryprogram router;
4664 (e) A message specified by the cannot_route_message router option.
4666 In these cases only, Exim does include the error details in bounce and
4667 warning messages. There are also a few cases where bland messages such
4668 as "unrouteable address" or "local delivery error" are given.
4670 PH/43 $value is now also set for the "else" part of a ${run expansion.
4672 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4673 being worked on, but at least Exim now implements the latest version to
4676 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4677 process to complete was treated as a delivery failure, a timeout while
4678 writing the message to the pipe was logged, but erroneously treated as a
4679 successful delivery. Such timeouts include transport filter timeouts. For
4680 consistency with the overall process timeout, these timeouts are now
4681 treated as errors, giving rise to delivery failures by default. However,
4682 there is now a new Boolean option for the pipe transport called
4683 timeout_defer, which, if set TRUE, converts the failures into defers for
4684 both kinds of timeout. A transport filter timeout is now identified in
4687 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4688 systems where "make" and "gmake" are different, calling "gmake" at top
4689 level broke things. I've arranged for the value of $(MAKE) to be passed
4690 from the Makefile to this script so that it can call the same version of
4694 A note about Exim versions 4.44 and 4.50
4695 ----------------------------------------
4697 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4698 changes of various kinds. As a consequence, a big documentation update was
4699 needed. This delayed the release for rather longer than seemed good, especially
4700 in the light of a couple of (minor) security issues. Therefore, the changes
4701 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4702 release. So 4.44 and 4.50 are in effect two different branches that both start
4705 I have left the 4.50 change log unchanged; it contains all the changes since
4706 4.43. The change log for 4.44 is below; many of its items are identical to
4707 those for 4.50. This seems to be the most sensible way to preserve the
4708 historical information.
4714 1. Minor wording change to the doc/README.SIEVE file.
4716 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4717 computation of the current number of files was incorrect.
4719 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4720 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4723 4. Give more explanation in the error message when the command for a transport
4724 filter fails to execute.
4726 5. There are several places where Exim runs a non-Exim command in a
4727 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4728 was being done only for the command run by the queryprogram router. It is
4729 now done for all such subprocesses. The other cases are: ${run, transport
4730 filters, and the commands run by the lmtp and pipe transports.
4732 6. Added CONFIGURE_GROUP build-time option.
4734 7. Some older OS have a limit of 256 on the maximum number of file
4735 descriptors. Exim was using setrlimit() to set 1000 as a large value
4736 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4737 systems. I've change it so that if it can't get 1000, it tries for 256.
4739 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4740 was an oversight, and furthermore, ever since the addition of extra
4741 controls (e.g. 4.43/32), the checks on when to allow different forms of
4742 "control" were broken. There should now be diagnostics for all cases when a
4743 control that does not make sense is encountered.
4745 9. Added the /retain_sender option to "control=submission".
4747 10. $recipients is now available in the predata ACL (oversight).
4749 11. Tidy the search cache before the fork to do a delivery from a message
4750 received from the command line. Otherwise the child will trigger a lookup
4751 failure and thereby defer the delivery if it tries to use (for example) a
4752 cached ldap connection that the parent has called unbind on.
4754 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4755 of $address_data from the recipient verification was clobbered by the
4756 sender verification.
4758 13. The value of address_data from a sender verification is now available in
4759 $sender_address_data in subsequent conditions in the ACL statement.
4761 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4763 15. Added a new option "connect=<time>" to callout options, to set a different
4766 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4767 was its contents. (It was OK if the option was not defined at all.)
4769 17. A "Completed" log line is now written for messages that are removed from
4770 the spool by the -Mrm option.
4772 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4773 information about exactly what failed.
4775 19. Added -dd to debug only the daemon process.
4777 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4778 handles timeouts, both on the server side and network timeouts. Renamed the
4779 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4781 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4782 It is now set to "smtps".
4784 22. $host_address is now set to the target address during the checking of
4785 ignore_target_hosts.
4787 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4788 being passed; this would have caused $sender_host_name to have been used if
4789 matching the list had actually called for a host name (not very likely,
4790 since this list is usually IP addresses). A host name is now passed as
4793 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4794 code that specifies a non-restarting handler (typically sigaction() in
4795 modern systems) in an attempt to fix a rare and obscure crash bug.
4797 25. Narrowed the window for a race in the daemon that could cause it to ignore
4798 SIGCHLD signals. This is not a major problem, because they are used only to
4799 wake it up if nothing else does.
4801 26. A malformed maildirsize file could cause Exim to calculate negative values
4802 for the mailbox size or file count. Odd effects could occur as a result.
4803 The maildirsize information is now recalculated if the size or filecount
4806 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4807 support for a long time. Removed HAVE_SYS_VFS_H.
4809 28. Installed the latest version of exipick from John Jetmore.
4811 29. In an address list, if the pattern was not a regular expression, an empty
4812 subject address (from a bounce message) matched only if the pattern was an
4813 empty string. Non-empty patterns were not even tested. This was the wrong
4814 because it is perfectly reasonable to use an empty address as part of a
4815 database query. An empty address is now tested by patterns that are
4816 lookups. However, all the other forms of pattern expect the subject to
4817 contain a local part and a domain, and therefore, for them, an empty
4818 address still always fails if the pattern is not itself empty.
4820 30. Exim went into a mad DNS loop when attempting to do a callout where the
4821 host was specified on an smtp transport, and looking it up yielded more
4822 than one IP address.
4824 31. Re-factored the code for checking spool and log partition space into a
4825 function that finds that data and another that does the check. The former
4826 is then used to implement four new variables: $spool_space, $log_space,
4827 $spool_inodes, and $log_inodes.
4829 32. The RFC2047 encoding function was originally intended for short strings
4830 such as real names; it was not keeping to the 75-character limit for
4831 encoded words that the RFC imposes. It now respects the limit, and
4832 generates multiple encoded words if necessary. To be on the safe side, I
4833 have increased the buffer size for the ${rfc2047: expansion operator from
4836 33. It is now permitted to omit both strings after an "if" condition; if the
4837 condition is true, the result is "true". As before, when the second string
4838 is omitted, a false condition yields an empty string. This makes it less
4839 cumbersome to write custom ACL and router conditions.
4841 34. Failure to deliver a bounce message always caused it to be frozen, even if
4842 there was an errors_to setting on the router. The errors_to setting is now
4845 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4846 canonical form (fully expanded) before being placed in
4847 $sender_host_address.
4849 36. The table in the code that translates DNS record types into text (T_A to
4850 "A" for instance) was missing entries for NS and CNAME. It is just possible
4851 that this could have caused confusion if both these types were looked up
4852 for the same domain, because the text type is used as part of Exim's
4853 per-process caching. But the chance of anyone hitting this buglet seems
4856 37. The dnsdb lookup has been extended in a number of ways.
4858 (1) There is a new type, "zns", which walks up the domain tree until it
4859 finds some nameserver records. It should be used with care.
4861 (2) There is a new type, "mxh", which is like "mx" except that it returns
4862 just the host names, not the priorities.
4864 (3) It is now possible to give a list of domains (or IP addresses) to be
4865 looked up. The behaviour when one of the lookups defers can be
4866 controlled by a keyword.
4868 (4) It is now possible to specify the separator character for use when
4869 multiple records are returned.
4871 38. The dnslists ACL condition has been extended: it is now possible to supply
4872 a list of IP addresses and/or domains to be looked up in a particular DNS
4875 39. Added log_selector=+queue_time_overall.
4877 40. When running the queue in the test harness, wait just a tad after forking a
4878 delivery process, to get repeatability of debugging output.
4880 41. Include certificate and key file names in error message when GnuTLS fails
4881 to set them up, because the GnuTLS error message doesn't include the name
4882 of the failing file when there is a problem reading it.
4884 42. Allow both -bf and -bF in the same test run.
4886 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4888 44. The "Exiscan patch" is now merged into the mainline Exim source.
4890 45. Sometimes the final signoff response after QUIT could fail to get
4891 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4892 < 0 before doing a fflush(). This bug looks as though it goes back to the
4893 introduction of TLS in release 3.20, but "sometimes" must have been rare
4894 because the tests only now provoked it.
4896 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4897 (this can affect the format of dates).
4899 47. exim_tidydb, when checking for the continued existence of a message for
4900 which it has found a message-specific retry record, was not finding
4901 messages that were in split spool directories. Consequently, it was
4902 deleting retry records that should have stayed in existence.
4904 48. Steve fixed some bugs in eximstats.
4906 49. The SPA authentication driver was not abandoning authentication and moving
4907 on to the next authenticator when an expansion was forced to fail,
4908 contradicting the general specification for all authenticators. Instead it
4909 was generating a temporary error. It now behaves as specified.
4911 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4912 (the order specifies the preference for clients). The order is now AES256,
4913 AES128, 3DES, ARCFOUR128.
4915 51. Small patch to Sieve code - explicitly set From: when generating an
4918 52. Exim crashed if a remote delivery caused a very long error message to be
4919 recorded - for instance if somebody sent an entire SpamAssassin report back
4920 as a large number of 550 error lines. This bug was coincidentally fixed by
4921 increasing the size of one of Exim's internal buffers (big_buffer) that
4922 happened as part of the Exiscan merge. However, to be on the safe side, I
4923 have made the code more robust (and fixed the comments that describe what
4926 53. Now that there can be additional text after "Completed" in log lines (if
4927 the queue_time_overall log selector is set), a one-byte patch to exigrep
4928 was needed to allow it to recognize "Completed" as not the last thing in
4931 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4932 patch that reportedly fixes this has been added. I am not expert enough to
4933 create a test for it. This is what the patch creator wrote:
4935 "I found a little strange behaviour of ldap code when working with
4936 Windows 2003 AD Domain, where users was placed in more than one
4937 Organization Units. When I tried to give exim partial DN, the exit code
4938 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4939 But simultaneously result of request was absolutely normal ldap result,
4940 so I produce this patch..."
4942 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4943 REFERENCE, so I have modified the code to exclude the patch when that macro
4946 55. Some experimental protocols are using DNS PTR records for new purposes. The
4947 keys for these records are domain names, not reversed IP addresses. The
4948 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4949 leaves it alone. Component reversal etc. now happens only for IP addresses.
4952 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4954 57. Double the size of the debug message buffer (to 2048) so that more of very
4955 long debug lines gets shown.
4957 58. The exicyclog utility now does better if the number of log files to keep
4958 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4960 59. Two changes related to the smtp_active_hostname option:
4962 (1) $smtp_active_hostname is now available as a variable.
4963 (2) The default for smtp_banner uses $smtp_active_hostname instead
4964 of $primary_hostname.
4966 60. The host_aton() function is supposed to be passed a string that is known
4967 to be a valid IP address. However, in the case of IPv6 addresses, it was
4968 not checking this. This is a hostage to fortune. Exim now panics and dies
4969 if the condition is not met. A case was found where this could be provoked
4970 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4971 components; fortuitously, this particular loophole had already been fixed
4972 by change 4.50/55 above.
4974 If there are any other similar loopholes, the new check in host_aton()
4975 itself should stop them being exploited. The report I received stated that
4976 data on the command line could provoke the exploit when Exim was running as
4977 exim, but did not say which command line option was involved. All I could
4978 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4979 running as the user.
4982 61. There was a buffer overflow vulnerability in the SPA authentication code
4983 (which came originally from the Samba project). I have added a test to the
4984 spa_base64_to_bits() function which I hope fixes it.
4987 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4988 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4990 63. The daemon start-up calls getloadavg() while still root for those OS that
4991 need the first call to be done as root, but it missed one case: when
4992 deliver_queue_load_max is set with deliver_drop_privilege. This is
4993 necessary for the benefit of the queue runner, because there is no re-exec
4994 when deliver_drop_privilege is set.
4996 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4997 This has been fixed.
4999 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5000 was in use, was not putting the data itself into the right store pool;
5001 consequently, it could be overwritten for a subsequent message in the same
5002 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5005 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5007 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5008 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5009 the function were treating the return as a boolean value, which happened to
5010 work because 0=false and not-0=true, but is not correct code.
5012 68. The host_aton() function was not handling scoped IPv6 addresses (those
5013 with, for example, "%eth0" on the end) correctly.
5015 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5016 negated items (that is, ~something) in unsigned ints. Some compilers
5017 apparently mutter when there is no cast.
5019 70. If an address verification called from an ACL failed, and did not produce a
5020 user-specific message (i.e. there was only a "system" message), nothing was
5021 put in $acl_verify_message. In this situation, it now puts the system
5024 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5025 session; change 4.31/43 added the unwanted input to the log line - except
5026 that it did not do this in the start of session case. It now does.
5028 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5029 This is wrong and can cause the other end to generate a synchronization
5030 error if it is another Exim or anything else that does the synchronization
5031 check. A QUIT command is no longer sent after a timeout.
5033 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5034 during host lookups.
5036 74. The defer_ok option of callout verification was not working if it was used
5037 when verifying addresses in header lines, that is, for this case:
5039 verify = header_sender/callout=defer_ok
5041 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5042 those file descriptors could be used for SMTP connections. If anything
5043 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5044 could be sent to the SMTP client, causing chaos. The daemon now opens
5045 stdin, stdout, and stderr to /dev/null when it puts itself into the
5048 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5049 log by default. The user can override this with suitable Perl magic.
5051 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5052 the log message when discard triggers, was not working for the DATA ACL or
5053 for the non-SMTP ACL.
5055 78. Error message wording change in sieve.c.
5057 79. If smtp_accept_max_per_host was set, the number of connections could be
5058 restricted to fewer than expected, because the daemon was trying to set up
5059 a new connection before checking whether the processes handling previous
5060 connections had finished. The check for completed processes is now done
5061 earlier. On busy systems, this bug wouldn't be noticed because something
5062 else would have woken the daemon, and it would have reaped the completed
5065 80. If a message was submitted locally by a user whose login name contained one
5066 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5067 It caused a spool format error. I have fixed the spool reading code. A
5068 related problem was that the "from" clause in the Received: line became
5069 illegal because of the space(s). It is now covered by ${quote_local_part.
5071 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5074 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5076 83. Patch from Sieve maintainer to fix unterminated string problem in
5077 "vacation" handling.
5079 84. Some minor changes to the Linux configuration files to help with other
5080 OS variants using glibc.
5082 85. One more patch for Sieve to update vacation handling to latest spec.
5085 ----------------------------------------------------
5086 See the note above about the 4.44 and 4.50 releases.
5087 ----------------------------------------------------
5093 1. Change 4.43/35 introduced a bug that caused file counts to be
5094 incorrectly computed when quota_filecount was set in an appendfile
5097 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5098 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5101 3. Give more explanation in the error message when the command for a transport
5102 filter fails to execute.
5104 4. There are several places where Exim runs a non-Exim command in a
5105 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5106 was being done only for the command run by the queryprogram router. It is
5107 now done for all such subprocesses. The other cases are: ${run, transport
5108 filters, and the commands run by the lmtp and pipe transports.
5110 5. Some older OS have a limit of 256 on the maximum number of file
5111 descriptors. Exim was using setrlimit() to set 1000 as a large value
5112 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5113 systems. I've change it so that if it can't get 1000, it tries for 256.
5115 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5116 was an oversight, and furthermore, ever since the addition of extra
5117 controls (e.g. 4.43/32), the checks on when to allow different forms of
5118 "control" were broken. There should now be diagnostics for all cases when a
5119 control that does not make sense is encountered.
5121 7. $recipients is now available in the predata ACL (oversight).
5123 8. Tidy the search cache before the fork to do a delivery from a message
5124 received from the command line. Otherwise the child will trigger a lookup
5125 failure and thereby defer the delivery if it tries to use (for example) a
5126 cached ldap connection that the parent has called unbind on.
5128 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5129 of $address_data from the recipient verification was clobbered by the
5130 sender verification.
5132 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5133 was its contents. (It was OK if the option was not defined at all.)
5135 11. A "Completed" log line is now written for messages that are removed from
5136 the spool by the -Mrm option.
5138 12. $host_address is now set to the target address during the checking of
5139 ignore_target_hosts.
5141 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5142 being passed; this would have caused $sender_host_name to have been used if
5143 matching the list had actually called for a host name (not very likely,
5144 since this list is usually IP addresses). A host name is now passed as
5147 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5148 code that specifies a non-restarting handler (typically sigaction() in
5149 modern systems) in an attempt to fix a rare and obscure crash bug.
5151 15. Narrowed the window for a race in the daemon that could cause it to ignore
5152 SIGCHLD signals. This is not a major problem, because they are used only to
5153 wake it up if nothing else does.
5155 16. A malformed maildirsize file could cause Exim to calculate negative values
5156 for the mailbox size or file count. Odd effects could occur as a result.
5157 The maildirsize information is now recalculated if the size or filecount
5160 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5161 support for a long time. Removed HAVE_SYS_VFS_H.
5163 18. Updated exipick to current release from John Jetmore.
5165 19. Allow an empty sender to be matched against a lookup in an address list.
5166 Previously the only cases considered were a regular expression, or an
5169 20. Exim went into a mad DNS lookup loop when doing a callout where the
5170 host was specified on the transport, if the DNS lookup yielded more than
5173 21. The RFC2047 encoding function was originally intended for short strings
5174 such as real names; it was not keeping to the 75-character limit for
5175 encoded words that the RFC imposes. It now respects the limit, and
5176 generates multiple encoded words if necessary. To be on the safe side, I
5177 have increased the buffer size for the ${rfc2047: expansion operator from
5180 22. Failure to deliver a bounce message always caused it to be frozen, even if
5181 there was an errors_to setting on the router. The errors_to setting is now
5184 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5185 canonical form (fully expanded) before being placed in
5186 $sender_host_address.
5188 24. Updated eximstats to version 1.33
5190 25. Include certificate and key file names in error message when GnuTLS fails
5191 to set them up, because the GnuTLS error message doesn't include the name
5192 of the failing file when there is a problem reading it.
5194 26. Expand error message when OpenSSL has problems setting up cert/key files.
5197 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5198 (this can affect the format of dates).
5200 28. exim_tidydb, when checking for the continued existence of a message for
5201 which it has found a message-specific retry record, was not finding
5202 messages that were in split spool directories. Consequently, it was
5203 deleting retry records that should have stayed in existence.
5205 29. eximstats updated to version 1.35
5206 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5207 1.35 - bugfix such that pie charts by volume are generated correctly
5209 30. The SPA authentication driver was not abandoning authentication and moving
5210 on to the next authenticator when an expansion was forced to fail,
5211 contradicting the general specification for all authenticators. Instead it
5212 was generating a temporary error. It now behaves as specified.
5214 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5215 (the order specifies the preference for clients). The order is now AES256,
5216 AES128, 3DES, ARCFOUR128.
5218 31. Small patch to Sieve code - explicitly set From: when generating an
5221 32. Exim crashed if a remote delivery caused a very long error message to be
5222 recorded - for instance if somebody sent an entire SpamAssassin report back
5223 as a large number of 550 error lines. This bug was coincidentally fixed by
5224 increasing the size of one of Exim's internal buffers (big_buffer) that
5225 happened as part of the Exiscan merge. However, to be on the safe side, I
5226 have made the code more robust (and fixed the comments that describe what
5229 33. Some experimental protocols are using DNS PTR records for new purposes. The
5230 keys for these records are domain names, not reversed IP addresses. The
5231 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5232 leaves it alone. Component reversal etc. now happens only for IP addresses.
5235 34. The host_aton() function is supposed to be passed a string that is known
5236 to be a valid IP address. However, in the case of IPv6 addresses, it was
5237 not checking this. This is a hostage to fortune. Exim now panics and dies
5238 if the condition is not met. A case was found where this could be provoked
5239 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5240 components; fortuitously, this particular loophole had already been fixed
5241 by change 4.50/55 or 4.44/33 above.
5243 If there are any other similar loopholes, the new check in host_aton()
5244 itself should stop them being exploited. The report I received stated that
5245 data on the command line could provoke the exploit when Exim was running as
5246 exim, but did not say which command line option was involved. All I could
5247 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5248 running as the user.
5251 35. There was a buffer overflow vulnerability in the SPA authentication code
5252 (which came originally from the Samba project). I have added a test to the
5253 spa_base64_to_bits() function which I hope fixes it.
5256 36. The daemon start-up calls getloadavg() while still root for those OS that
5257 need the first call to be done as root, but it missed one case: when
5258 deliver_queue_load_max is set with deliver_drop_privilege. This is
5259 necessary for the benefit of the queue runner, because there is no re-exec
5260 when deliver_drop_privilege is set.
5262 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5263 was in use, was not putting the data itself into the right store pool;
5264 consequently, it could be overwritten for a subsequent message in the same
5265 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5268 38. Sometimes the final signoff response after QUIT could fail to get
5269 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5270 < 0 before doing a fflush(). This bug looks as though it goes back to the
5271 introduction of TLS in release 3.20, but "sometimes" must have been rare
5272 because the tests only now provoked it.
5278 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5279 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5280 Now it returns an integer. A number of calls were still expecting a T/F
5281 return. Fortuitously, in all cases, the tests worked in OK situations,
5282 which is the norm. However, things would have gone wrong on any write
5283 failures on the smtp file descriptor. This function is used when sending
5284 messages over SMTP and also when doing verify callouts.
5286 2. When Exim is called to do synchronous delivery of a locally submitted
5287 message (the -odf or -odi options), it no longer closes stderr before doing
5290 3. Implemented the mua_wrapper option.
5292 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5294 5. Implemented the functions header_remove(), header_testname(),
5295 header_add_at_position(), and receive_remove_recipient(), and exported them
5298 6. If an ACL "warn" statement specified the addition of headers, Exim already
5299 inserted X-ACL-Warn: at the start if there was no header name. However, it
5300 was not making this test for the second and subsequent header lines if
5301 there were newlines in the string. This meant that an invalid header could
5302 be inserted if Exim was badly configured.
5304 7. Allow an ACL "warn" statement to add header lines at the start or after all
5305 the Received: headers, as well as at the end.
5307 8. Added the rcpt_4xx retry error code.
5309 9. Added postmaster_mailfrom=xxx to callout verification option.
5311 10. Added mailfrom=xxxx to the callout verification option, for verify=
5314 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5315 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5317 12. Inserted some casts to stop certain compilers warning when using pointer
5318 differences as field lengths or precisions in printf-type calls (mostly
5319 affecting debugging statements).
5321 13. Added optional readline() support for -be (dynamically loaded).
5323 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5324 same clock tick as a message's arrival, so that its received time was the
5325 same as the "first fail" time on the retry record, and that message
5326 remained on the queue past the ultimate address timeout, every queue runner
5327 would try a delivery (because it was past the ultimate address timeout) but
5328 after another failure, the ultimate address timeout, which should have then
5329 bounced the address, did not kick in. This was a "< instead of <=" error;
5330 in most cases the first failure would have been in the next clock tick
5331 after the received time, and all would be well.
5333 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5334 being recognized when the domain list was tested by the match_domain
5335 condition in an expansion string.
5337 16. Added the ${str2b64: operator.
5339 17. Exim was always calling setrlimit() to set a large limit for the number of
5340 processes, without checking whether the existing limit was already
5341 adequate. (It did check for the limit on file descriptors.) Furthermore,
5342 errors from getrlimit() and setrlimit() were being ignored. Now they are
5343 logged to the main and panic logs, but Exim does carry on, to try to do its
5344 job under whatever limits there are.
5346 18. Imported PCRE 5.0.
5348 19. Trivial typo in log message " temporarily refused connection" (the leading
5351 20. If the log selector return_path_on_delivery was set and an address was
5352 redirected to /dev/null, the delivery process crashed because it assumed
5353 that a return path would always be set for a "successful" delivery. In this
5354 case, the whole delivery is bypassed as an optimization, and therefore no
5357 21. Internal re-arrangement: the function for sending a challenge and reading
5358 a response while authentication was assuming a zero-terminated challenge
5359 string. It's now changed to take a pointer and a length, to allow for
5360 binary data in such strings.
5362 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5364 23. Exim was not respecting finduser_retries when seeking the login of the
5365 uid under which it was called; it was always trying 10 times. (The default
5366 setting of finduser_retries is zero.) Also, it was sleeping after the final
5367 failure, which is pointless.
5369 24. Implemented tls_on_connect_ports.
5371 25. Implemented acl_smtp_predata.
5373 26. If the domain in control=submission is set empty, Exim assumes that the
5374 authenticated id is a complete email address when it generates From: or
5375 Sender: header lines.
5377 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5378 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5379 chown and chgrp in /bin and hostname in /usr/bin.
5381 28. Exim was keeping the "process log" file open after each use, just as it
5382 does for the main log. This opens the possibility of it remaining open for
5383 long periods when the USR1 signal hits a daemon. Occasional processlog
5384 errors were reported, that could have been caused by this. Anyway, it seems
5385 much more sensible not to leave this file open at all, so that is what now
5388 29. The long-running daemon process does not normally write to the log once it
5389 has entered its main loop, and it closes the log before doing so. This is
5390 so that log files can straightforwardly be renamed and moved. However,
5391 there are a couple of unusual error situations where the daemon does write
5392 log entries, and I had neglected to close the log afterwards.
5394 30. The text of an SMTP error response that was received during a remote
5395 delivery was being truncated at 512 bytes. This is too short for some of
5396 the long messages that one sometimes sees. I've increased the limit to
5399 31. It is now possible to make retry rules that apply only when a message has a
5400 specific sender, in particular, an empty sender.
5402 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5403 it possible to be selective about when SMTP synchronization is enforced.
5405 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5407 32. Implemented hosts_connection_nolog.
5409 33. Added an ACL for QUIT.
5411 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5414 35. Added mailbox_size and mailbox_filecount to appendfile.
5416 36. Added control = no_multiline_responses to ACLs.
5418 37. There was a bug in the logic of the code that waits for the clock to tick
5419 in the case where the clock went backwards by a substantial amount such
5420 that the microsecond fraction of "now" was more than the microsecond
5421 fraction of "then" (but the whole seconds number was less).
5423 38. Added support for the libradius Radius client library this is found on
5424 FreeBSD (previously only the radiusclient library was supported).
5430 1. When certain lookups returned multiple values in the form name=value, the
5431 quoting of the values was not always being done properly. Specifically:
5432 (a) If the value started with a double quote, but contained no whitespace,
5434 (b) If the value contained whitespace other than a space character (i.e.
5435 tabs or newlines or carriage returns) it was not quoted.
5436 This fix has been applied to the mysql and pgsql lookups by writing a
5437 separate quoting function and calling it from the lookup code. The fix
5438 should probably also be applied to nisplus, ibase and oracle lookups, but
5439 since I cannot test any of those, I have not disturbed their existing code.
5441 2. A hit in the callout cache for a specific address caused a log line with no
5442 reason for rejecting RCPT. Now it says "Previous (cached) callout
5443 verification failure".
5445 3. There was an off-by-one bug in the queryprogram router. An over-long
5446 return line was truncated at 256 instead of 255 characters, thereby
5447 overflowing its buffer with the terminating zero. As well as fixing this, I
5448 have increased the buffer size to 1024 (and made a note to document this).
5450 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5451 when Exim is waiting for an SMTP response from a remote server, Exim
5452 restarts its select() call on the socket, thereby resetting its timeout.
5453 This is not a problem when such interrupts are rare. Somebody set up a cron
5454 job to run exiwhat every 2 minutes, which is less than the normal select()
5455 timeout (5 or 10 minutes). This meant that the select() timeout never
5456 kicked in because it was always reset. I have fixed this by comparing the
5457 time when an interrupt arrives with the time at the start of the first call
5458 to select(). If more time than the timeout has elapsed, the interrupt is
5459 treated as a timeout.
5461 5. Some internal re-factoring in preparation for the addition of Sieve
5462 extensions (by MH). In particular, the "personal" test is moved to a
5463 separate function, and given an option for scanning Cc: and Bcc: (which is
5464 not set for Exim filters).
5466 6. When Exim created an email address using the login of the caller as the
5467 local part (e.g. when creating a From: or Sender: header line), it was not
5468 quoting the local part when it contained special characters such as @.
5470 7. Installed new OpenBSD configuration files.
5472 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5473 try to make them clearer.
5475 9. Callout options, other than the timeout value, were being ignored when
5476 verifying sender addresses in header lines. For example, when using
5478 verify = header_sender/callout=no_cache
5480 the cache was (incorrectly) being used.
5482 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5483 only the Cygwin environment.
5485 11. When return_path_on_delivery was set as a log selector, if different remote
5486 addresses in the same message used different return paths and parallel
5487 remote delivery occurred, the wrong values would sometimes be logged.
5488 (Whenever a remote delivery process finished, the return path value from
5489 the most recently started remote delivery process was logged.)
5491 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5492 lines when AUTH and/or TLS are in use. This is the "received protocol"
5493 field. Exim used to use "asmtp" for authenticated SMTP, without any
5494 indication (in the protocol name) for TLS use. Now it follows the RFC and
5495 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5496 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5497 names appear in log lines as well as in Received: header lines.
5499 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5500 extensions, and comparison tests, and to fix some bugs.
5502 14. Changes to the "personal" filter test:
5504 (1) The test was buggy in that it was just doing the equivalent of
5505 "contains" tests on header lines. For example, if a user's address was
5506 anne@some.where, the "personal" test would incorrectly be true for
5508 To: susanne@some.where
5510 This test is now done by extracting each address from the header in turn,
5511 and checking the entire address. Other tests that are part of "personal"
5512 are now done using regular expressions (for example, to check local parts
5513 of addresses in From: header lines).
5515 (2) The list of non-personal local parts in From: addresses has been
5516 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5517 taken from the Sieve specification recommendations.
5519 (3) If the message contains any header line starting with "List-" it is
5520 treated as non-personal.
5522 (4) The test for "circular" in the Subject: header line has been removed
5523 because it now seems ill-conceived.
5525 15. Minor typos in src/EDITME comments corrected.
5527 16. Installed latest exipick from John Jetmore.
5529 17. If headers_add on a router specified a text string that was too long for
5530 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5531 of string_sprintf() is now avoided.
5533 18. $message_body_size was not set (it was always zero) when running the DATA
5534 ACL and the local_scan() function.
5536 19. For the "mail" command in an Exim filter, no default was being set for
5537 the once_repeat time, causing a random time value to be used if "once" was
5538 specified. (If the value happened to be <= 0, no repeat happened.) The
5539 default is now 0s, meaning "never repeat". The "vacation" command was OK
5540 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5541 (I found it when inspecting the code).
5543 20. There is now an overall timeout for performing a callout verification. It
5544 defaults to 4 times the callout timeout, which applies to individual SMTP
5545 commands during the callout. The overall timeout applies when there is more
5546 than one host that can be tried. The timeout is checked before trying the
5547 next host. This prevents very long delays if there are a large number of
5548 hosts and all are timing out (e.g. when the network connections are timing
5549 out). The value of the overall timeout can be changed by specifying an
5550 additional sub-option for "callout", called "maxwait". For example:
5552 verify = sender/callout=5s,maxwait=20s
5554 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5555 to the end before writing, but this should make it even safer).
5557 22. Exim was forgetting that it had advertised PIPELINING for the second and
5558 subsequent messages on an SMTP connection. It was also not resetting its
5559 memory on STARTTLS and an internal HELO.
5561 23. When Exim logs an SMTP synchronization error within a session, it now
5562 records whether PIPELINING has been advertised or not.
5564 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5565 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5566 rather than long int.
5568 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5570 26. Added the never_mail option to autoreply.
5576 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5577 crash if the getsockname() call failed; this can happen if a connection is
5578 closed very soon after it is established. The problem was simply in the
5579 order in which certain operations were done, causing Exim to try to write
5580 to the SMTP stream before it had set up the file descriptor. The bug has
5581 been fixed by making things happen in the correct order.
5587 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5588 before the connection was closed, thus losing the rejection response.
5590 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5591 some early Solaris releases, but causes trouble in current releases where
5592 socklen_t is defined.
5594 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5597 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5600 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5601 that releasing the top of it at the end releases what was used for sub-
5602 expansions (unless the block got too big). However, discard this block if
5603 the first thing is a variable or header, so that we can use its block when
5604 it is dynamic (useful for very large $message_headers, for example).
5606 6. Lookups now cache *every* query, not just the most recent. A new, separate
5607 store pool is used for this. It can be recovered when all lookup caches are
5608 flushed. Lookups now release memory at the end of their result strings.
5609 This has involved some general refactoring of the lookup sources.
5611 7. Some code has been added to the store_xxx() functions to reduce the amount
5612 of flapping under certain conditions.
5614 8. log_incoming_interface used to affect only the <= reception log lines. Now
5615 it causes the local interface and port to be added to several more SMTP log
5616 lines, for example "SMTP connection from", and rejection lines.
5618 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5620 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5622 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5623 could be overwritten at the end of the current message (or the start of a
5624 new message if it was set in a HELO ACL). The value is now preserved for
5625 the duration of the SMTP connection.
5627 12. If a transport had a headers_rewrite setting, and a matching header line
5628 contained an unqualified address, that address was qualified, even if it
5629 did not match any rewriting rules. The underlying bug was that the values
5630 of the flags that permit the existence of unqualified sender and recipient
5631 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5632 non-local messages, and by -bnq for local messages) were not being
5633 preserved with the message after it was received.
5635 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5636 "next input=" as part of the text comprising the host identity instead of
5637 the correct text. The code was using the same buffer for two different
5638 strings. However, depending on which order the printing function evaluated
5639 its arguments, the bug did not always show up. Under Linux, for example, my
5640 test suite worked just fine.
5642 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5643 doesn't work with some older versions of Perl. It has been changed to "my",
5644 which in any case is probably the better facility to use.
5646 15. A really picky compiler found some instances of statements for creating
5647 error messages that either had too many or two few arguments for the format
5650 16. The size of the buffer for calls to the DNS resolver has been increased
5651 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5652 for addresses that have a lot of PTR records. This alleviates a problem; it
5653 does not fully solve it.
5655 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5656 buffer now truncates the list and logs the incident, which is the same
5657 action as happens when Exim is looking up a host name and its aliases.
5658 Previously in this situation something unpredictable would happen;
5659 sometimes it was "internal error: store_reset failed".
5661 18. If a server dropped the connection unexpectedly when an Exim client was
5662 using GnuTLS and trying to read a response, the client delivery process
5663 crashed while trying to generate an error log message.
5665 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5666 string, for example:
5668 warn message = H1: something\nH2: something
5670 the text was added as a single header line from Exim's point of view
5671 though it ended up OK in the delivered message. However, searching for the
5672 second and subsequent header lines using $h_h2: did not work. This has been
5673 fixed. Similarly, if a system filter added multiple headers in this way,
5674 the routers could not see them.
5676 20. Expanded the error message when iplsearch is called with an invalid key to
5677 suggest using net-iplsearch in a host list.
5679 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5680 ACLs are no longer actually imposed (and a message to that effect is
5683 22. If a "gecos" field in a passwd entry contained escaped characters, in
5684 particular, if it contained a \" sequence, Exim got it wrong when building
5685 a From: or a Sender: header from that name. A second bug also caused
5686 incorrect handling when an unquoted " was present following a character
5687 that needed quoting.
5689 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5690 was not being matched caselessly.
5692 24. Arranged for all hyphens in the exim.8 source to be escaped with
5695 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5696 properly. Recipient callout cache records were still being keyed to include
5697 the sender, even when use_sender was set false. This led to far more
5698 callouts that were necessary. The sender is no longer included in the key
5699 when use_sender is false.
5701 26. Added "control = submission" modifier to ACLs.
5703 27. Added the ${base62d: operator to decode base 62 numbers.
5705 28. dnsdb lookups can now access SRV records.
5707 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5708 the configuration file.
5710 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5711 (-v) mode. This makes the output for a verbose queue run more intelligible.
5713 31. Added a use_postmaster feature to recipient callouts.
5715 32. Added the $body_zerocount variable, containing the number of binary zero
5716 bytes in the message body.
5718 33. The time of last modification of the "new" subdirectory is now used as the
5719 "mailbox time last read" when there is a quota error for a maildir
5722 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5724 35. Added +ignore_unknown as a special item in host lists.
5726 36. Code for decoding IPv6 addresses in host lists is now included, even if
5727 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5728 address was recognized as an IP address, but was then not correctly decoded
5729 into binary, causing unexpected and incorrect effects when compared with
5736 1. Very minor rewording of debugging text in manualroute to say "list of
5737 hosts" instead of "hostlist".
5739 2. If verify=header_syntax was set, and a header line with an unqualified
5740 address (no domain) and a large number of spaces between the end of the
5741 name and the colon was received, the reception process suffered a buffer
5742 overflow, and (when I tested it) crashed. This was caused by some obsolete
5743 code that should have been removed. The fix is to remove it!
5745 3. When running in the test harness, delay a bit after writing a bounce
5746 message to get a bit more predictability in the log output.
5748 4. Added a call to search_tidyup() just before forking a reception process. In
5749 theory, someone could use a lookup in the expansion of smtp_accept_max_
5750 per_host which, without the tidyup, could leave open a database connection.
5752 5. Added the variables $recipient_data and $sender_data which get set from a
5753 lookup success in an ACL "recipients" or "senders" condition, or a router
5754 "senders" option, similar to $domain_data and $local_part_data.
5756 6. Moved the writing of debug_print from before to after the "senders" test
5759 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5760 problems for message scanning, either using a data ACL, or using
5761 local_scan() because the Received: header was not generated till after they
5762 were called (in order to set the time as the time of reception completion).
5763 I have revised the way this works. The header is now generated after the
5764 body is received, but before the ACL or local_scan() are called. After they
5765 are run, the timestamp in the header is updated.
5771 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5772 before starting a queue runner without re-exec. This happened only when
5773 deliver_drop_privilege was set or when the Exim user was set to root. The
5774 effect of the bug was that timeouts during subsequent deliveries caused
5775 crashes instead of being properly handled. The handler is now left at its
5776 default (and expected) setting.
5778 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5779 message, again when deliver_drop_privilege is set or Exim is run as root.
5780 The bug described in (1) was not present in this case, but the tidying up
5781 of the other signals was missing. I have made the two cases consistent.
5783 3. The ignore_target_hosts setting on a manualroute router was being ignored
5784 for hosts that were looked up using the /MX notation.
5786 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5789 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5790 operated on the sender address. After changing the $sender_address to <>
5791 for the sender address verify, Exim was re-instated it as the original
5792 (before rewriting) address, but remembering that it had rewritten it, so it
5793 wasn't rewriting it again. This bug also had the effect of breaking the
5794 sender address verification caching when the sender address was rewritten.
5796 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5797 This has been changed so that if the ip literal address matches
5798 ignore_target_hosts, the router declines.
5800 7. Added expansion conditions match_domain, match_address, and match_local_
5801 part (NOT match_host).
5803 8. The placeholder for the Received: header didn't have a length field set.
5805 9. Added code to Exim itself and to exim_lock to test for a specific race
5806 condition that could lead to file corruption when using MBX delivery. The
5807 issue is with the lockfile that is created in /tmp. If this file is removed
5808 after a process has opened it but before that process has acquired a lock,
5809 there is the potential for a second process to recreate the file and also
5810 acquire a lock. This could lead to two Exim processes writing to the file
5811 at the same time. The added code performs the same test as UW imapd; it
5812 checks after acquiring the lock that its file descriptor still refers to
5813 the same named file.
5815 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5816 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5819 11. Added the smtp_active_hostname option. If used, this will typically be made
5820 to depend on the incoming interface address. Because $interface_address is
5821 not set up until the daemon has forked a reception process, error responses
5822 that can happen earlier (such as "too many connections") no longer contain
5825 12. If an expansion in a condition on a "warn" statement fails because a lookup
5826 defers, the "warn" statement is abandoned, and the next ACL statement is
5827 processed. Previously this caused the whole ACL to be aborted.
5829 13. Added the iplsearch lookup type.
5831 14. Added ident_timeout as a log selector.
5833 15. Added tls_certificate_verified as a log selector.
5835 16. Added a global option tls_require_ciphers (compare the smtp transport
5836 option of the same name). This controls incoming TLS connections.
5838 17. I finally figured out how to make tls_require_ciphers do a similar thing
5839 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5840 before starting the TLS session.
5842 18. Tabs are now shown as \t in -bP output.
5844 19. If the log selector return_path_on_delivery was set, Exim crashed when
5845 bouncing a message because it had too many Received: header lines.
5847 20. If two routers both had headers_remove settings, and the first one included
5848 a superfluous trailing colon, the final name in the first list and the
5849 first name in the second list were incorrectly joined into one item (with a
5850 colon in the middle).
5856 1. Added -C and -D options to the exinext utility, mainly to make it easier
5857 to include in the automated testing, but these could be helpful when
5858 multiple configurations are in use.
5860 2. The exinext utility was not formatting the output nicely when there was
5861 an alternate port involved in the retry record key, nor when there was a
5862 message id as well (for retries that were specific to a specific message
5863 and a specific host). It was also confused by IPv6 addresses, because of
5864 the additional colons they contain. I have fixed the IPv4 problem, and
5865 patched it up to do a reasonable job for IPv6.
5867 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5868 delivery, the log line now contains "pipelined" if PIPELINING was used.
5870 4. An SMTP transport process used to panic and die if the bind() call to set
5871 an explicit outgoing interface failed. This has been changed; it is now
5872 treated in the same way as a connect() failure.
5874 5. A reference to $sender_host_name in the part of a conditional expansion
5875 that was being skipped was still causing a DNS lookup. This no longer
5878 6. The def: expansion condition was not recognizing references to header lines
5879 that used bh_ and bheader_.
5881 7. Added the _cache feature to named lists.
5883 8. The code for checking quota_filecount in the appendfile transport was
5884 allowing one more file than it should have been.
5886 9. For compatibility with Sendmail, the command line option
5894 and sets the incoming protocol and host name (for trusted callers). The
5895 host name and its colon can be omitted when only the protocol is to be set.
5896 Note the Exim already has two private options, -pd and -ps, that refer to
5897 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5898 "s", but I don't think that's a major issue.
5900 10. A number of refactoring changes to the code, none of which should affect
5903 (a) The number of logging options was getting close to filling up the
5904 32-bit word that was used as a bit map. I have split them into two classes:
5905 those that are passed in the argument to log_write(), and those that are
5906 only ever tested independently outside of that function. These are now in
5907 separate 32-bit words, so there is plenty of room for expansion again.
5908 There is no change in the user interface or the logging behaviour.
5910 (b) When building, for example, log lines, the code previously used a
5911 macro that called string_cat() twice, in order to add two strings. This is
5912 not really sufficiently general. Furthermore, there was one instance where
5913 it was actually wrong because one of the argument was used twice, and in
5914 one call a function was used. (As it happened, calling the function twice
5915 did not affect the overall behaviour.) The macro has been replaced by a
5916 function that can join an arbitrary number of extra strings onto a growing
5919 (c) The code for expansion conditions now uses a table and a binary chop
5920 instead of a serial search (which was left over from when there were very
5921 few conditions). Also, it now recognizes conditions like "pam" even when
5922 the relevant support is not compiled in: a suitably worded error message is
5923 given if an attempt is made to use such a condition.
5925 11. Added ${time_interval:xxxxx}.
5927 12. A bug was causing one of the ddress fields not to be passed back correctly
5928 from remote delivery subprocesses. The field in question was not being
5929 subsequently used, so this caused to problems in practice.
5931 13. Added new log selectors queue_time and deliver_time.
5933 14. Might have fixed a bug in maildirsizefile handling that threw up
5934 "unexpected character" debug warnings, and recalculated the data
5935 unnecessarily. In any case, I expanded the warning message to give more
5938 15. Added the message "Restricted characters in address" to the statements in
5939 the default ACL that block characters like @ and % in local parts.
5941 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5942 Three changes have been made:
5944 (a) There was a serious bug; a negative response to MAIL caused the whole
5945 recipient domain to be cached as invalid, thereby blocking all messages
5946 to all local parts at the same domain, from all senders. This bug has
5947 been fixed. The domain is no longer cached after a negative response to
5948 MAIL if the sender used is not empty.
5950 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5953 (c) A new callout option, "use_sender" has been added for people who want
5954 the modified behaviour.
5960 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5963 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5964 indeed breaks things for older releases.
5966 3. Added additional logging to the case where there is a problem reading data
5967 from a filter that is running in a subprocess using a pipe, in order to
5968 try to track down a specific problem.
5970 4. Testing facility fudge: when running in the test harness and attempting
5971 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5972 getting "No route to host". Convert this to a timeout.
5974 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5977 6. Some OS don't have socklen_t but use size_t instead. This affects the
5978 fifth argument of getsockopt() amongst other things. This is now
5979 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5980 can be set for individual OS. I have set it for SunOS5, OSF1, and
5981 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5982 some earlier ones do not.
5984 7. Change 4.30/15 was not doing the test caselessly.
5986 8. The standard form for an IPv6 address literal was being rejected by address
5987 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5988 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5989 this, as well as the form without the "IPv6" on the front (but only when
5990 address literals are enabled, of course).
5992 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5994 10. Exim crashed if a message with an empty sender address specified by -f
5995 encountered a router with an errors_to setting. This could be provoked only
5996 by a command such as
6000 where an empty string was supplied; "<>" did not hit this bug.
6002 11. Installed PCRE release 4.5.
6004 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6005 remained set. It is now erased.
6007 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6008 times from message ids (which are base 36 rather than the normal 62).
6010 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6011 were being counted as actual protocol errors, and logged if the log
6012 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6013 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6014 and DATA following a set of rejected RCPTs do not count as protocol errors.
6015 In other words, Exim assumes they were pipelined, though this may not
6016 actually be the case. Of course, in all cases the client gets an
6017 appropriate error code.
6019 15. If a lookup fails in an ACL condition, a message about the failure may
6020 be available; it is used if testing the ACL cannot continue, because most
6021 such messages specify what the cause of the deferral is. However, some
6022 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6023 that caused an old message to be retained and used if a later statement
6024 caused a defer, replacing the real cause of the deferral.
6026 16. If an IP address had so many PTR records that the DNS lookup buffer
6027 was not large enough to hold them, Exim could crash while trying to process
6028 the truncated data. It now detects and logs this case.
6030 17. Further to 4.21/58, another change has been made: if (and only if) the
6031 first line of a message (the first header line) ends with CRLF, a bare LF
6032 in a subsequent header line has a space inserted after it, so as not to
6033 terminate the header.
6035 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6036 unnecessarily, used atoi() instead of strtol(), and didn't check the
6037 termination when getting file sizes from file names by regex.
6039 19. Completely re-implemented the support for maildirsize files, in the light
6040 of a number of problems with the previous contributed implementation
6041 (4.30/29). In particular:
6043 . If the quota is zero, the maildirsize file is maintained, but no quota is
6046 . If the maildir directory does not exist, it is created before any attempt
6047 to write a maildirsize file.
6049 . The quota value in the file is just a cache; if the quota is changed in
6050 the transport, the new value overrides.
6052 . A regular expression is available for excluding directories from the
6055 20. The autoreply transport checks the characters in options that define the
6056 message's headers; it allows continued headers, but it was checking with
6057 isspace() after an embedded newline instead of explicitly looking for a
6060 21. If all the "regular" hosts to which an address was routed had passed their
6061 expiry times, and had not reached their retry times, the address was
6062 bounced, even if fallback hosts were defined. Now Exim should go on to try
6065 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6066 equivalent code in the SMTP transport. Some hosts send humungous responses
6067 to HELO/EHLO, more than 1024 it seems.
6069 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6070 gives compiler warnings in some environments. I've now done it "properly",
6073 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6074 (because of problems with the built-in one) was declared to return uschar *
6075 instead of char *, causing compiler failure.
6077 25. Fixed a file descriptor leak when processing alias/forward files.
6079 26. Fixed a minor format string issue in dbfn.c.
6081 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6083 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6084 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6087 29. When a sender address is verified, it is cached, to save repeating the test
6088 when there is more than one recipient in a message. However, when the
6089 verification involves a callout, it is possible for different callout
6090 options to be set for different recipients. It is too complicated to keep
6091 track of this in the cache, so now Exim always runs a verification when a
6092 callout is required, relying on the callout cache for the optimization.
6093 The overhead is duplication of the address routing, but this should not be
6096 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6097 to be verified with callout=postmaster, and the main callout worked but the
6098 postmaster check failed, the verification correctly failed. However, if a
6099 subsequent RCPT command asked for sender verification *without* the
6100 postmaster check, incorrect caching caused this verification also to fail,
6103 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6104 it was not caching the DNS options (qualify_single, search_parents) that
6105 were used when the lookup failed. A subsequent lookup with different
6106 options therefore always gave the same answer, though there were cases
6107 where it should not have. (Example: a "domains = !$mx_any" option on a
6108 dnslookup router: the "domains" option is always processed without any
6109 widening, but the router might have qualify_single set.) Now Exim uses the
6110 cached value only when the same options are set.
6112 32. Added John Jetmore's "exipick" utility to the distribution.
6114 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6115 than a timeout (e.g. a certificate is required, and is not provided), an
6116 Exim server now closes the connection immediately. Previously it waited for
6117 the client to close - but if the client is SSL, it seems that they each
6118 wait for each other, leading to a delay before one of them times out.
6120 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6121 maintained 0.8.x compatibility because I don't think many are using it, and
6122 it is clearly obsolete.
6124 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6127 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6128 client certificate was expired. A simple patch fixes this, though I don't
6129 understand the full logic of why the verify callback is called multiple
6132 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6133 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6134 which causes problems with some clients (such as the Certicom SSL Plus
6135 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6136 disables the coutermeasure allowing Eudora to connect."
6138 38. Exim was not checking that a write() to a log file succeeded. This could
6139 lead to Bad Things if a log got too big, in particular if it hit a file
6140 size limit. Exim now panics and dies if it cannot write to a log file, just
6141 as it does if it cannot open a log file.
6143 39. Modified OS/Makefile-Linux so that it now contains
6145 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6147 The two -D definitions ensure that Exim is compiled with large file
6148 support, which makes it possible to handle log files that are bigger than
6151 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6152 instance) a domain was checked against a named list that involved a lookup,
6153 causing $domain_data to be set, then another domain was checked against the
6154 same list, then the first domain was re-checked, the value of $domain_data
6155 after the final check could be wrong. In particular, if the second check
6156 failed, it could be set empty. This bug probably also applied to
6159 41. The strip_trailing_dot option was not being applied to the address given
6160 with the -f command-line option.
6162 42. The code for reading a message's header from the spool was incrementing
6163 $received_count, but never initializing it. This meant that the value was
6164 incorrect (doubled) while delivering a message in the same process in which
6165 it was received. In the most common configuration of Exim, this never
6166 happens - a fresh exec is done - but it can happen when
6167 deliver_drop_privilege is set.
6169 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6170 it now includes up to 150 characters of the unexpected data in the log
6173 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6174 and building data strings. The size of both of these buffers was 10 000
6175 bytes - far larger than anybody would *ever* want, thought I. Needless to
6176 say, somebody hit the limit. I have increased the maximum line length to
6177 20 000 and the maximum data length of concatenated lines to 100 000. I have
6178 also fixed two bugs, because there was no checking on these buffers. Tsk,
6179 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6180 buffer is too small.
6182 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6183 lsearch lookups. Now it does.
6185 46. When parsing a route_list item in a manualroute router, a fixed-length
6186 buffer was used for the list of hosts. I made this 1024 bytes long,
6187 thinking that nobody would ever have a list of hosts that long. Wrong.
6188 Somebody had a whole pile of complicated expansion conditions, and the
6189 string was silently truncated, leading to an expansion error. It turns out
6190 that it is easier to change to an unlimited length (owing to other changes
6191 that have happened since this code was originally written) than to build
6192 structure for giving a limitation error. The length of the item that
6193 expands into the list of hosts is now unlimited.
6195 47. The lsearch lookup could not handle data where the length of text line was
6196 more than 4095 characters. Such lines were truncated, leading to shortened
6197 data being returned. It should now handle lines of any length.
6199 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6200 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6203 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6204 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6205 no longer happens in comment lines. A list of replacements is now placed
6206 at the head of all of the source files, except those whose only change is
6207 to replace PERL_COMMAND in the very first #! line.
6209 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6210 messages on the queue, with a bottom-up merge sort, using code contributed
6211 by Michael Haardt. This should make operations like -bp somewhat faster on
6212 large queues. It won't affect queue runners, except when queue_run_in_order
6215 51. Installed eximstats 1.31 in the distribution.
6217 52. Added support for SRV lookups to the dnslookup router.
6219 53. If an ACL referred to $message_body or $message_body_end, the value was not
6220 reset for any messages that followed in the same SMTP session.
6222 54. The store-handling optimization for building very long strings was not
6223 differentiating between the different store pools. I don't think this
6224 actually made any difference in practice, but I've tidied it.
6226 55. While running the routers to verify a sender address, $sender_address
6227 was still set to the sender address. This is wrong, because when routing to
6228 send a bounce to the sender, it would be empty. Therefore, I have changed
6229 it so that, while verifying a sender address, $sender_address is set to <>.
6230 (There is no change to what happens when verifying a recipient address.)
6232 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6233 target A or AAAA records (if not already returned) without resetting the
6234 qualify_single or search_parents options of the DNS resolver. These are
6235 inappropriate in this case because the targets of MX and SRV records must
6236 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6237 target that, when qualified, matched something in the local domain. These
6238 two options are now turned off when doing these lookups.
6240 57. It seems that at least some releases of Reiserfs (which does not have the
6241 concept of a fixed number of inodes) returns zero and not -1 for the
6242 number of available inodes. This interacted badly with check_spool_inodes,
6243 which assumed that -1 was the "no such thing" setting. What I have done is
6244 to check that the total number of inodes is greater than zero before doing
6245 the test of how many are available.
6247 58. When a "warn" ACL statement has a log_message modifier, the message is
6248 remembered, and not repeated. This is to avoid a lot of repetition when a
6249 message has many recipients that cause the same warning to be written.
6250 However, Exim was preserving the list of already written lines for an
6251 entire SMTP session, which doesn't seem right. The memory is now reset if a
6252 new message is started.
6254 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6255 debugging output unless log_rewrite was also set.
6257 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6258 of (int)(handle) when we know that handle contains (void *)(-1).
6260 61. The Exim daemon panic-logs an error return when it closes the incoming
6261 connection. However "connection reset by peer" seems to be common, and
6262 isn't really an error worthy of noting specially, so that particular error
6265 62. When Exim is trying to find all the local interfaces, it used to panic and
6266 die if the ioctl to get the interface flags failed. However, it seems that
6267 on at least one OS (Solaris 9) it is possible to have an interface that is
6268 included in the list of interfaces, but for which you get a failure error
6269 for this call. This happens when the interface is not "plumbed" into a
6270 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6271 failure of the "get flags" call assumes that the interface is down.
6273 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6274 makes life easier for people who are doing arithmetic on fields extracted
6275 from dates, where you often get leading zeros that should not be
6276 interpreted as octal.
6278 64. Added qualify_domain to the redirect router, to override the global
6281 65. If a pathologically long header line contained very many addresses (the
6282 report of this problem mentioned 10 000) and each of them was rewritten,
6283 Exim could use up a very large amount of memory. (It kept on making new
6284 copies of the header line as it rewrote, and never released the old ones.)
6285 At the expense of a bit more processing, the header rewriting function has
6286 been changed so that it no longer eats memory in this way.
6288 66. The generation of the Received: header has been moved from the time that a
6289 message starts to be received, to the time that it finishes. The timestamp
6290 in the Received: header should now be very close to that of the <= log
6291 line. There are two side-effects of this change:
6293 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6294 logged header lines no longer include the local Received: line, because
6295 it has not yet been created. The same applies to a copy of the message
6296 that is returned to a non-SMTP sender when a message is rejected.
6298 (b) When a filter file is tested using -bf, no additional Received: header
6299 is added to the test message. After some thought, I decided that this
6302 This change does not affect the value of $received_for. It is still set
6303 after address rewriting, but before local_scan() is called.
6305 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6307 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6308 gave an unhelpful panic error message, and a defer error. I have managed to
6309 change this behaviour so that it now rejects any supplied certificate,
6310 which seems right, as the list of acceptable certificates is empty.
6312 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6313 gave an unhelpful defer error. I have not managed to make this reject any
6314 supplied certificates, but the error message it gives is "no certificate
6315 supplied", which is not helpful.
6317 70. exigrep's output now also includes lines that are not associated with any
6318 message, but which match the given pattern. Implemented by a patch from
6319 Martin Sluka, which also tidied up the Perl a bit.
6321 71. Recipient callout verification, like sender verification, was using <> in
6322 the MAIL FROM command. This isn't really the right thing, since the actual
6323 sender may affect whether the remote host accepts the recipient or not. I
6324 have changed it to use the actual sender in the callout; this means that
6325 the cache record is now keyed on a recipient/sender pair, not just the
6326 recipient address. There doesn't seem to be a real danger of callout loops,
6327 since a callout by the remote host to check the sender would use <>.
6328 [SEE ABOVE: changed after hitting problems.]
6330 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6331 temporary errors. However, in the case of such a code being given after
6332 the end of a data transmission (i.e. after ".") Exim was failing to write
6333 a retry record for the message. (Yes, there was some broken host that was
6334 actually sending 8xx at this point.)
6336 73. An unknown lookup type in a host list could cause Exim to panic-die when
6337 the list was checked. (An example that provoked this was putting <; in the
6338 middle of a list instead of at the start.) If this happened during a DATA
6339 ACL check, a -D file could be left lying around. This kind of configuration
6340 error no longer causes Exim to die; instead it causes a defer error. The
6341 incident is still logged to the main and panic logs.
6343 74. Buglet left over from Exim 3 conversion. The message "too many messages
6344 in one connection" was written to the rejectlog but not the mainlog, except
6345 when address rewriting (yes!) was being logged.
6347 75. Added write_rejectlog option.
6349 76. When a system filter was run not as root (that is, when system_filter_user
6350 was set), the values of the $n variables were not being returned to the
6351 main process; thus, they were not subsequently available in the $sn
6354 77. Added +return_path_on_delivery log selector.
6356 78. A connection timeout was being treated differently from recipients deferred
6357 when testing hosts_max_try with a message that was older than the host's
6358 retry timeout. (The host should not be counted, thus allowing all hosts to
6359 be tried at least once before bouncing.) This may have been the cause of an
6360 occasionally reported bug whereby a message would remain on the queue
6361 longer than the retry timeout, but would be bounced if a delivery was
6362 forced. I say "may" because I never totally pinned down the problem;
6363 setting up timeout/retry tests is difficult. See also the next item.
6365 79. The ultimate address timeout was not being applied to errors that involved
6366 a combination of host plus message (for example, a timeout on a MAIL
6367 command). When an address resolved to a number of possible hosts, and they
6368 were not all tried for each delivery (e.g. because of hosts_max_try), a
6369 message could remain on the queue longer than the retry timeout.
6371 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6374 81. Fixed an obscure SMTP outgoing bug which required at least the following
6375 conditions: (a) there was another message waiting for the same server;
6376 (b) the server returned 5xx to all RCPT commands in the first message so
6377 that the message was not completed; (c) the server dropped the connection
6378 or gave a negative response to the RSET that Exim sends to abort the
6379 transaction. The observed case was a dropped connection after DATA that had
6380 been sent in pipelining mode. That is, the server had advertised PIPELINING
6381 but was not implementing it correctly. The effect of the bug was incorrect
6382 behaviour, such as trying another host, and this could lead to a crash.
6388 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6389 and daemon.c were passed as pointers to ints; they should have been
6390 pointers to socklen_t variables (which are typically unsigned ints).
6392 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6395 3. Fixed a really odd bug that affected only the testing scheme; patching a
6396 certain fixed string in the binary changed the value of another string that
6397 happened to be identical to the end of the original first string.
6399 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6400 name", it returns that address as the IP address. On some operating
6401 systems (e.g. Solaris), it also passes back the IP address string as the
6402 "host name". However, on others (e.g. Linux), it passes back an empty
6403 string. Exim wasn't checking for this, and was changing the host name to an
6404 empty string, assuming it had been canonicalized.
6406 5. Although rare, it is permitted to have more than one PTR record for a given
6407 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6408 all the names associated with an address, because they do in Solaris.
6409 However, it seems that they do not in Linux for data that comes from the
6410 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6411 I found this out when I moved to a new Linux workstation and tried to run
6412 the Exim test suite.
6414 To get round this problem I have changed the code so that it now does its
6415 own call to the DNS to look up PTR records when searching for a host name.
6416 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6417 addresses that are only in /etc/hosts are still found.
6419 This behaviour is, however, controlled by an option called host_lookup_
6420 order, which defaults to "bydns:byaddr". If people want to use the other
6421 order, or indeed, just use one or the other means of lookup, they can
6422 specify it in this variable.
6424 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6425 some operating systems, this comes back from gethostbyaddr() as an empty
6426 string, and this is what Exim used to test for. However, it seems that in
6427 other systems, "." is yielded. Exim now tests for this case too.
6429 7. The values of check_spool_space and check_log_space are now held internally
6430 as a number of kilobytes instead of an absolute number of bytes. If a
6431 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6432 kilobyte. This means that much larger values can be stored.
6434 8. Exim monitor: an attempt to get the action menu when not actually pointing
6435 at a message produces an empty menu entitled "No message selected". This
6436 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6437 no entries in it ("Shell widget menu has zero width and/or height"). So I
6438 have added a single, blank menu entry in this case.
6440 9. Added ${quote_local_part.
6442 10. MIME decoding is now applied to the contents of Subject: header lines when
6445 11. Now that a reference to $sender_host_address automatically causes a reverse
6446 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6447 host lookup before query-style lookups in lists that might use this
6448 variable. This has therefore been abolished, and the "net-" prefix is no
6449 longer necessary for query-style lookups.
6451 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6452 have been a typo for LFLAGS, so it has been changed.
6454 13. The install script calls Exim with "-C /dev/null" in order to find the
6455 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6456 to be output. However, since Exim outputs its version number before the
6457 error, it didn't break the script. It just looked ugly. I fixed this by
6458 always allowing "-C /dev/null" if the caller is root.
6460 14. Ignore overlarge ACL variable number when reading spool file - insurance
6461 against a later release with more variables having written the file.
6463 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6464 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6465 this, as well as the form without the "IPv6" on the front.
6467 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6468 OS/Makefile-Darwin file.
6470 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6471 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6473 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6474 variable. However, some OS use ints for inodes. I've added cast to long int
6475 to get rid of the compiler warning.
6477 19. I had forgotten to lock out "/../" in configuration file names when
6478 ALT_CONFIG_PREFIX was set.
6480 20. Routers used for verification do not need to specify transports. However,
6481 if such a router generated a host list, and callout was configured, Exim
6482 crashed, because it could not find a port number from the (non-existent)
6483 transport. It now assumes port 25 in this circumstance.
6485 21. Added the -t option to exigrep.
6487 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6488 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6489 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6490 them is set, LOOKUP_LSEARCH is forced.
6492 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6494 24. Added sender and host information to the "rejected by local_scan()" log
6495 line; previously there was no indication of these.
6497 25. Added .include_if_exists.
6499 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6500 Linux. It turns out that not all file systems support this. Apparently some
6501 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6502 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6503 not supported on the file descriptor, is now ignored when Exim is trying to
6504 sync a directory. This applies only to Linux.
6506 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6508 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6509 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6512 29. Added support for maildirsize files from supplied patch (modified a bit).
6514 30. The use of :fail: followed by an empty string could lead Exim to respond to
6515 sender verification failures with (e.g.):
6517 550 Verification failed for <xxx>
6518 550 Sender verify failed
6520 where the first response line was missing the '-' that indicates it is not
6521 the final line of the response.
6523 31. The loop for finding the name of the user that called Exim had a hardwired
6524 limit of 10; it now uses the value of finduser_retries, which is used for
6525 all other user lookups.
6527 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6530 33. Exim was neglecting to zero errno before one call of strtol() when
6531 expanding a string and expecting an integer value. On some systems this
6532 resulted in spurious "integer overflow" errors. Also, it was casting the
6533 result into an int without checking.
6535 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6536 did not work. The code looks as if it has *never* worked, though it appears
6537 to have been documented since at least release 1.62. I have made it work.
6539 35. The "timeout_DNS" error in retry rules, also documented since at least
6540 1.62, also never worked. As it isn't clear exactly what this means, and
6541 clearly it isn't a major issue, I have abolished the feature by treating it
6542 as "timeout", and writing a warning to the main and panic logs.
6544 36. The display of retry rules for -brt wasn't always showing the error code
6547 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6548 timeout_connect_A, timeout_connect_MX.
6550 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6551 to the empty sender.
6553 39. The daemon was not analysing the content of -oX till after it had closed
6554 stderr and disconnected from the controlling terminal. This meant that any
6555 syntax errors were only noted on the panic log, and the return code from
6556 the command was 0. By re-arranging the code a little, I've made the
6557 decoding happen first, so such errors now appear on stderr, and the return
6558 code is 1. However, the actual setting up of the sockets still happens in
6559 the disconnected process, so errors there are still only recorded on the
6562 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6563 connections (as happens on some IP stacks) was logged at start up time as
6564 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6565 it from "IPv6 and IPv4", which means that two separate sockets are being
6568 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6569 says whether AF_INET or AF_INET6 was passed as an argument.
6571 42. Exiwhat output was messed up when time zones were included in log
6574 43. Exiwhat now gives more information about the daemon's listening ports,
6575 and whether -tls-on-connect was used.
6577 44. The "port" option of the smtp transport is now expanded.
6579 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6580 silently ignored. Now an error message is written to the main and panic
6583 46. There's a new ACL modifier called "logwrite" which writes to a log file
6584 as soon as it is encountered.
6586 47. Added $local_user_uid and $local_user_gid at routing time.
6588 48. Exim crashed when trying to verify a sender address that was being
6591 49. Exim was recognizing only a space character after ".include". It now also
6592 recognizes a tab character.
6594 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6595 extracting the relevant information from the specification. The man page no
6596 longer contains scrambled data for the -d option, and I've added a section
6597 at the front about calling Exim under different names.
6599 51. Added "extra_headers" argument to the "mail" command in filter files.
6601 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6604 53. Installed eximstats 1.29.
6606 54. Added transport_filter_timeout as a generic transport option.
6608 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6609 Cc: header lines. This was required by RFC 822, but it not required by RFC
6612 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6613 incoming messages that did not have them. Now it does so only if the
6614 message originates locally, that is, if there is no associated remote host
6615 address. When Resent- header lines are present, this applies to the Resent-
6616 lines rather than the non-Resent- lines.
6618 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6619 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6621 58. Messages for configuration errors now include the name of the main
6622 configuration file - useful now that there may be more than one file in a
6623 list (.included file names were always shown).
6625 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6626 for those rare installations that do not start the daemon as root or run it
6627 setuid root. I've cut out the call to initgroups() if the daemon is not
6630 60. The Exim user and group can now be bound into the binary as text strings
6631 that are looked up at the start of Exim's processing.
6633 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6635 62. Added $mailstore_basename variable.
6637 63. Installed patch to sieve.c from Michael Haardt.
6639 64. When Exim failed to open the panic log after failing to open the main log,
6640 the original message it was trying to log was written to stderr and debug
6641 output, but if they were not available (the usual case in production), it
6642 was lost. Now it is written to syslog before the two lines that record the
6643 failures to open the logs.
6645 65. Users' Exim filters run in subprocesses under the user's uid. It is
6646 possible for a "deliver" command or an alias in a "personal" command to
6647 provoke an address rewrite. If logging of address rewriting is configured,
6648 this fails because the process is not running as root or exim. There may be
6649 a better way of dealing with this, but for the moment (because 4.30 needs
6650 to be released), I have disabled address rewrite logging when running a
6651 filter in a non-root, non-exim process.
6657 1. The buildconfig auxiliary program wasn't quoting the value set for
6658 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6659 not defined. This bug was masked in 4.22 by the effect that was fixed in
6662 2. Some messages that were rejected after a message id was allocated were
6663 shown as "incomplete" by exigrep. It no longer does this for messages that
6664 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6666 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6667 have allow_domain_literals set, the ID did not get logged in the <= line.
6668 Domain literals are now always recognized in Message-ID: header lines.
6670 4. The first argument for a ${extract expansion item is the key name or field
6671 number. Leading and trailing spaces in this item were not being ignored,
6672 causing some misleading effects.
6674 5. When deliver_drop_privilege was set, single queue runner processes started
6675 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6676 same command in the process it spins off) were not dropping privilege.
6678 6. When the daemon running as "exim" started a queue runner, it always
6679 re-executed Exim in the spun-off process. This is a waste of effort when
6680 deliver_drop_privilege is set. The new process now just calls the
6681 queue-runner function directly.
6687 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6690 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6691 ignored. Though the use of -oP was forcing the writing of a pid file, it
6692 was always written to the default place.
6694 3. If the message "no IP address found for host xxxx" is generated during
6695 incoming verification, it is now followed by identification of the incoming
6696 connection (so you can more easily find what provoked it).
6698 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6700 5. Added some features to "harden" Exim a bit more against certain attacks:
6702 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6703 be put in Local/Makefile. This is like the never_users runtime option,
6704 but it cannot be overridden. The default setting is "root".
6706 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6707 prefix string with which any file named in a -C command line option
6710 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6711 is retained for -C and -D only if the caller of Exim is root. Without
6712 it, the exim user may also use -C and -D and retain privilege.
6714 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6715 command line option is disabled.
6717 6. Macro names set by the -D option must start with an upper case letter, just
6718 like macro names defined in the configuration file.
6720 7. Added "dereference=" facility to LDAP.
6722 8. Two instances of the typo "uknown" in the source files are fixed.
6724 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6725 the Configure-Makefile script screwed up while processing it.
6727 10. Incorporated PCRE 4.4.
6729 11. The SMTP synchronization check was not operating right at the start of an
6730 SMTP session. For example, it could not catch a HELO sent before the client
6731 waited for the greeting. There is now a check for outstanding input at the
6732 point when the greeting is written. Because of the duplex, asynchronous
6733 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6734 way, but not yet received, when the check is performed.
6736 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6737 on TCP/IP sockets, because this apparently causes some broken clients to
6740 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6741 unchanged) from the Cygwin maintainer.
6743 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6744 was supported instead of testing for mbx. Effectively a typo.
6746 15. The spa authenticator server code was not checking that the input it
6747 received was valid base64.
6749 16. The debug output line for the "set" modifier in ACLs was not showing the
6750 name of the variable that was being set.
6752 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6754 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6755 lookup on its own. Something else was needed to trigger the lookup. For
6756 example, a match in host_lookup or the need for a host name in a host list.
6757 Now, if $sender_host_name is referenced and the host name has not yet been
6758 looked up, a lookup is performed. If the lookup fails, the variable remains
6759 empty, and $host_lookup_failed is set to "1".
6761 19. Added "eqi" as a case-independent comparison operator.
6763 20. The saslauthd authentication condition could segfault if neither service
6764 nor realm was specified.
6766 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6767 error message that was logged was misleading, and incoming SMTP
6768 connections were dropped. The message is now more accurate, and temporary
6769 errors are given to SMTP connections.
6771 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6772 (except RSET) with a 421 error, until QUIT is received. However, it was
6773 failing to send a response to QUIT.
6775 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6776 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6777 ACL is now run after the helo_try_verify_hosts code.
6779 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6780 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6781 case-independent, so other case variants are also recognized). Apparently
6782 some systems use these upper case variants.
6784 25. If more than two messages were waiting for the same host, and a transport
6785 filter was specified for the transport, Exim sent two messages over the
6786 same TCP/IP connection, and then failed with "socket operation on non-
6787 socket" when it tried to send the third.
6789 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6791 27. The extern definition of crypt16() in expand.c was not being excluded when
6792 the OS had its own crypt16() function.
6794 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6795 as a preferred synonym for return_size_limit, both as an option and as an
6798 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6800 30. Changed the default configuration ACL to relax the local part checking rule
6801 for addresses that are not in any local domains. For these addresses,
6802 slashes and pipe symbols are allowed within local parts, but the sequence
6803 /../ is explicitly forbidden.
6805 31. SPA server authentication was not clearing the challenge buffer before
6808 32. log_message in a "warn" ACL statement was writing to the reject log as
6809 well as to the main log, which contradicts the documentation and doesn't
6810 seem right (because no rejection is happening). So I have stopped it.
6812 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6813 However, I am unable to do any testing of this.
6815 34. Fixed an infelicity in the appendfile transport. When checking directories
6816 for a mailbox, to see if any needed to be created, it was accidentally
6817 using path names with one or more superfluous leading slashes; tracing
6818 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6820 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6821 contents are added to the log line that is written for every discarded
6822 recipient. (Previously a log_message setting was ignored.)
6824 36. The ${quote: operator now quotes the string if it is empty.
6826 37. The install script runs exim in order to find its version number. If for
6827 some reason other than non-existence or emptiness, which it checks, it
6828 could not run './exim', it was installing it with an empty version number,
6829 i.e. as "exim-". This error state is now caught, and the installation is
6832 38. An argument was missing from the function that creates an error message
6833 when Exim fails to connect to the socket for saslauthd authentication.
6834 This could cause Exim to crash, or give a corrupted message.
6836 39. Added isip, isip4, and isip6 to ${if conditions.
6838 40. The ACL variables $acl_xx are now saved with the message, and can be
6839 accessed later in routers, transports, and filters.
6841 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6842 strings in the file are not string-expanded.
6844 42. If a MAIL command specified a SIZE value that was too large to fit into an
6845 int variable, the check against message_size_limit failed. Such values are
6846 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6847 day this will have to be increased, but I don't think I want to be around
6848 when emails are that large.
6855 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6856 iconv() is not standard in FreeBSD.
6858 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6859 IPv6 enabled. The observed symptom was a segmentation fault on return from
6860 the function os_common_find_running_interfaces() in src/os.c.
6862 3. In the check_special_case() function in daemon.c I had used "errno" as an
6863 argument name, which causes warnings on some systems. This was basically a
6864 typo, since it was named "eno" in the comments!
6866 4. The code that waits for the clock to tick (at a resolution of some fraction
6867 of a second) so as to ensure message-id uniqueness was always waiting for
6868 at least one whole tick, when it could have waited for less. [This is
6869 almost certainly not relevant at current processor speeds, where it is
6870 unlikely to ever wait at all. But we try to future-proof.]
6872 5. The function that sleeps for a time interval that includes fractions of a
6873 second contained a race. It did not block SIGALRM between setting the
6874 timer, and suspending (a couple of lines later). If the interval was short
6875 and the sigsuspend() was delayed until after it had expired, the suspension
6876 never ended. On busy systems this could lead to processes getting stuck for
6879 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6880 process, before it forks any delivery processes. The open lookup caching
6881 mechanism meant that the open file or database connection was passed into
6882 the delivery process. The problem was that delivery processes always tidy
6883 up cached lookup data. This could cause a problem for the next delivery
6884 process started by the queue runner, because the external queue runner
6885 process does not know about the closure. So the next delivery process
6886 still has data in the lookup cache. In the case of a file lookup, there was
6887 no problem because closing a file descriptor in a subprocess doesn't affect
6888 the parent. However, if the lookup was caching a connection to a database,
6889 the connection was closed, and the second delivery process was likely to
6890 see errors such as "PGSQL: query failed: server closed the connection
6891 unexpectedly". The problem has been fixed by closing all cached lookups
6892 in a queue runner before running a delivery process.
6894 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6895 seem to have the "const" qualifier which it has on other OS. I've
6898 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6899 *of the same type* (client or server) with the same public name that an
6900 error should be diagnosed.
6902 9. When Exim looked up a host name for an IP address, but failed to find the
6903 original IP address when looking up the host name (a safety check), it
6904 output the message "<ip address> does not match any IP for NULL", which was
6905 confusing, to say the least. The bug was that the host name should have
6906 appeared instead of "NULL".
6908 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6909 user that is built into the binary, and the -C or -D options is used, root
6910 privilege is dropped before the configuration file is read. In addition,
6911 logging is switched to stderr instead of the normal log files. If the
6912 configuration then re-defines the Exim user, the unprivileged environment
6913 is probably not what is expected, so Exim logs a panic warning message (but
6916 However, if deliver_drop_privilege is set, the unprivileged state may well
6917 be exactly what is intended, so the warning has been cut out in that case,
6918 and Exim is allowed to try to write to its normal log files.
6924 1. smtp_return_error_details was not giving details for temporary sender
6925 or receiver verification errors.
6927 2. Diagnose a configuration error if two authenticators have the same public
6930 3. Exim used not to create the message log file for a message until the first
6931 delivery attempt. This could be confusing when incoming messages were held
6932 for policy or load reasons. The message log file is now created at the time
6933 the message is received, and an initial "Received" line is written to it.
6935 4. The automatically generated man page for command line options had a minor
6936 bug that caused no ill effects; however, a more serious problem was that
6937 the procedure for building the man page automatically didn't always
6938 operate. Consequently, release 4.20 contains an out-of-date version. This
6939 shouldn't happen again.
6941 5. When building Exim with embedded Perl support, the script that builds the
6942 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6943 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6945 6. The freeze_tell option was not being used for messages that were frozen on
6946 arrival, either by an ACL or by local_scan().
6948 7. Added the smtp_incomplete_transaction log selector.
6950 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6951 was accepting AUTH without a new EHLO.
6953 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6954 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6957 10. Logging of TCP/IP connections (when configured) now happens in the main
6958 daemon process instead of the child process, so that the TCP/IP connection
6959 count is more accurate (but it can never be perfect).
6961 11. The use of "drop" in a nested ACL was not being handled correctly in the
6962 outer ACL. Now, if condition failure induced by the nested "drop" causes
6963 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6964 or "require"), the connection is dropped.
6966 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6967 that yield "discard" can now be used with an "accept" or a "discard" verb,
6968 but an error is generated for any others (because I can't see a useful way
6969 to define what should happen).
6971 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6972 are now processed in the same way as lines in the Exim configuration file.
6973 In particular, continuation lines are supported.
6975 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6977 15. Added -ti meaning -t -i.
6979 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6980 domains, and warn by logging if others are found.
6982 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6983 returned from GIFCONF when Exim is trying to find the list of interfaces on
6984 a host. The code in os.c has been modified to copy each ifreq to an aligned
6985 structure in all cases.
6987 Also, in some cases, the returned ifreq's were being copied to a 'struct
6988 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6989 means the last couple of bytes of an IPv6 address could be chopped if the
6990 ifreq contained only a normal sockaddr (14 bytes storage).
6992 18. Named domain lists were not supported in the hosts_treat_as_local option.
6993 An entry such as +xxxx was not recognized, and was treated as a literal
6996 19. Ensure that header lines added by a DATA ACL are included in the reject log
6997 if the ACL subsequently rejects the message.
6999 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7000 MD5 (which is deprecated).
7002 21. When testing a filter file using -bf, Exim was writing a message when it
7003 took the sender from a "From " line in the message, but it was not doing so
7004 when it took $return_path from a Return-Path: header line. It now does.
7006 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7007 with a valid header line field name (a series of printing characters
7008 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7010 23. Changed "disc" in the source to "disk" to conform to the documentation and
7011 the book and for uniformity.
7013 24. Ignore Sendmail's -Ooption=value command line item.
7015 25. When execve() failed while trying to run a command in a pipe transport,
7016 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7017 could be confused with a return value of 69 from the command itself. This
7018 has been changed to 127, the value the shell returns if it is asked to run
7019 a non-existent command. The wording for the related log line suggests a
7020 non-existent command as the problem.
7022 26. If received_header_text expands to an empty string, do not add a Received:
7023 header line to the message. (Well, it adds a token one on the spool, but
7024 marks it "old" so that it doesn't get used or transmitted.)
7026 27. Installed eximstats 1.28 (addition of -nt option).
7028 28. There was no check for failure on the call to getsockname() in the daemon
7029 code. This can fail if there is a shortage of resources on the system, with
7030 ENOMEM, for example. A temporary error is now given on failure.
7032 29. Contrary to the C standard, it seems that in some environments, the
7033 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7034 program. Exim now does this explicitly; it affects the formatting of
7035 timestamps using strftime().
7037 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7038 complaints. I've now initialized all the variables, to avoid this.
7040 32. Header lines added by a system filter were not being "seen" during
7041 transport-time rewrites.
7043 33. The info_callback() function passed to OpenSSL is set up with type void
7044 (*)(SSL *, int, int), as described somewhere. However, when calling the
7045 function (actually a macro) that sets it up, the type void(*)() is
7046 expected. I've put in a cast to prevent warnings from picky compilers.
7048 34. If a DNS black list lookup found a CNAME record, but there were no A
7049 records associated with the domain it pointed at, Exim crashed.
7051 35. If a DNS black list lookup returned more than one A record, Exim ignored
7052 all but the first. It now scans all returned addresses if a particular IP
7053 value is being sought. In this situation, the contents of the
7054 $dnslist_value variable are a list of all the addresses, separated by a
7057 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7058 to accept a host name and all its aliases if the forward lookup for any of
7059 them yielded the IP address of the incoming connection. Now it accepts only
7060 those names whose forward lookup yields the correct IP address. Any other
7061 names are discarded. This closes a loophole whereby a rogue DNS
7062 administrator could create reverse DNS records to break through a
7063 wildcarded host restriction in an ACL.
7065 37. If a user filter or a system filter that ran in a subprocess used any of
7066 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7067 the wrong values were passed to the pipe command ($thisaddress had the
7068 value of $0, $0 had the value of $1, etc). This bug was introduced by
7069 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7071 38. Improved the line breaking for long SMTP error messages from ACLs.
7072 Previously, if there was no break point between 40 and 75 characters, Exim
7073 left the rest of the message alone. Two changes have been made: (a) I've
7074 reduced the minimum length to 35 characters; (b) if it can't find a break
7075 point between 35 and 75 characters, it looks ahead and uses the first one
7076 that it finds. This may give the occasional overlong line, but at least the
7077 remaining text gets split now.
7079 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7080 file descriptors might be low, and that setting 1000 would always raise it.
7081 It turns out that in some environments, the limit is already over 1000 and
7082 that lowering it causes trouble. So now Exim takes care not to decrease it.
7084 40. When delivering a message, the value of $return_path is set to $sender_
7085 address at the start of routing (routers may change the value). By an
7086 oversight, this default was not being set up when an address was tested by
7087 -bt or -bv, which affected the outcome if any router or filter referred to
7090 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7091 log only when "message" or "log_message" are set. However, if one of the
7092 conditions was an address verification, or a call to a nested ACL, the
7093 messages generated by the underlying test were being passed through. This
7094 no longer happens. The underlying message is available in $acl_verify_
7095 message for both "message" and "log_message" expansions, so it can be
7096 passed through if needed.
7098 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7099 new expansion $bh_ to give the encoded byte string without charset
7100 translation. Translation happens only if iconv() is available; HAVE_ICONV
7101 indicates this at build time. HEADERS_CHARSET gives the charset to
7102 translate to; headers_charset can change it in the configuration, and
7103 "headers charset" can change it in an individual filter file.
7105 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7106 that creates RFC 2047 encoded "words" labels them as that charset instead
7107 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7108 expansion operator; (ii) when Exim creates a From: line for a local
7109 message; (iii) when a header line is rewritten to include a "phrase" part.
7111 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7112 buggy, causing it to skip the first lines of messages whose message ID
7113 ended in 'D'. This would not have bitten before Exim release 4.14, because
7114 message IDs were unlikely to end in 'D' before then. The effect was to have
7115 incorrect size information for certain domains.
7117 45. #include "config.h" was missing at the start of the crypt16.c module. This
7118 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7121 46. If there was a timeout during a "random" callout check, Exim treated it as
7122 a failure of the random address, and carried on sending RSET and the real
7123 address. If the delay was just some slowness somewhere, the response to the
7124 original RCPT would be taken as a response to RSET and so on, causing
7125 mayhem of various kinds.
7127 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7128 when I implemented it. It didn't allow for the fact that some option values
7129 may legitimately be negative (e.g. size_addition), and it didn't even do
7130 the right test for positive values.
7132 48. Domain names in DNS records are case-independent. Exim always looks them up
7133 in lower case. Some resolvers return domain names in exactly the case they
7134 appear in the zone file, that is, they may contain uppercase letters. Not
7135 all resolvers do this - some return always lower case. Exim was treating a
7136 change of case by a resolver as a change of domain, similar to a widening
7137 of a domain abbreviation. This triggered its re-routing code and so it was
7138 trying to route what was effectively the same domain again. This normally
7139 caused routing to fail (because the router wouldn't handle the domain
7140 twice). Now Exim checks for this case specially, and just changes the
7141 casing of the domain that it ultimately uses when it transmits the message
7144 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7147 50. If a filter generated a file delivery with a non-absolute name (possible if
7148 no home directory exists for the router), the forbid_file option was not
7151 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7152 the existing equality matching.
7154 52. Exim was using ints instead of ino_t variables in some places where it was
7155 dealing with inode numbers.
7157 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7158 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7159 called TMPDIR, and if it finds it is different, it changes its value.
7161 54. The smtp_printf() function is now made available to local_scan() so
7162 additional output lines can be written before returning. There is also an
7163 smtp_fflush() function to enable the detection of a dropped connection.
7164 The variables smtp_input and smtp_batched_input are exported to
7167 55. Changed the default runtime configuration: the message "Unknown user"
7168 has been removed from the ACL, and instead placed on the localuser router,
7169 using the cannot_route_message feature. This means that any verification
7170 failures that generate their own messages won't get overridden. Similarly,
7171 the "Unrouteable address" message that was in the ACL for unverifiable
7172 relay addresses has also been removed.
7174 56. Added hosts_avoid_esmtp to the smtp transport.
7176 57. The exicyclog script was not checking for the esoteric option
7177 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7178 will work only if exicyclog is run under the appropriate euid.
7180 58. Following a discussion on the list, the rules by which Exim recognises line
7181 endings on incoming messages have been changed. The -dropcr and drop_cr
7182 options are now no-ops, retained only for backwards compatibility. The
7183 following line terminators are recognized: LF CRLF CR. However, special
7184 processing applies to CR:
7186 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7187 nor a local message in the state where . is a terminator.
7189 (ii) If a bare CR is encountered in a header line, an extra space is added
7190 after the line terminator so as not to end the header. The reasoning
7191 behind this is that bare CRs in header lines are most likely either
7192 to be mistakes, or people trying to play silly games.
7194 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7195 was being incorrectly given as 18 bytes larger than it should have been.
7196 This is a VOB (very old bug).
7198 60. This may never have affected anything current, but just in case it has:
7199 When the local host is found other than at the start of a list of hosts,
7200 the local host, those with the same MX, and any that follow, are discarded.
7201 When the list in question was part of a longer list of hosts, the following
7202 hosts (not currently being processed) were also being discarded. This no
7203 longer happens. I'm not sure if this situation could ever has previously
7206 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7209 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7210 header. This is something that is recommended in a new Internet Draft, and
7211 is something that is documented as being done by Sendmail. There are two
7212 possible values. For messages generated by the autoreply transport, Exim
7215 Auto-Submitted: auto-replied
7217 whereas for all other generated messages (e.g. bounces) it adds
7219 Auto-Submitted: auto-generated
7221 63. The "personal" condition in filters now includes a test for the
7222 Auto-Submitted: header. If it contains the string "auto-" the message it
7223 not considered personal.
7225 64. Added rcpt_include_affixes as a generic transport option.
7227 65. Added queue_only_override (default true).
7229 66. Added the syslog_duplication option.
7231 67. If what should have been the first header line of a message consisted of
7232 a space followed by a colon, Exim was mis-interpreting it as a header line.
7233 It isn't of course - it is syntactically invalid and should therefore be
7234 treated as the start of the message body. The misbehaviour could have
7235 caused a number of strange effects, including loss of data in subsequent
7236 header lines, and spool format errors.
7238 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7239 client host had authenticated. This control can now be exercised by an ACL
7240 for more flexibility.
7242 69. By default, callouts do not happen when testing with -bh. There is now a
7243 variant, -bhc, which does actually run the callout code, including
7244 consulting and updating the callout cache.
7246 70. Added support for saslauthd authentication, courtesy of Alexander
7249 71. If statvfs() failed on the spool or log directories while checking their
7250 size for availability, Exim confusingly gave the error "space shortage".
7251 Furthermore, in debugging mode it crashed with a floating point exception.
7252 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7253 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7254 serious problem, Exim now writes to the main and panic logs when this
7255 happens, with details of the failure. It then refuses to accept the
7256 incoming message, giving the message "spool directory problem" or "log
7257 directory problem" with a 421 code for SMTP messages.
7259 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7260 0, 1, and 2 exist, because some OS complain for execs without them (see
7261 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7262 descriptors. However, the code omitted to check that the open succeeded,
7263 causing mysterious errors if for some reason the permissions on /dev/null
7264 got screwed. Now Exim writes a message to the main and panic logs, and
7265 bombs out if it can't open /dev/null.
7267 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7268 interact so that it is all more flexible. It is supposed to remain
7269 backwards compatible. Also added extra_local_interfaces.
7271 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7272 to bomb out with an assertion failure - to the client this appears as a
7273 connection drop. This problem occurs in the part of the code that was taken
7274 from the Samba project. Fortunately, the assertion is in a very simple
7275 function, so I have fixed this by reproducing the function inline in the
7276 one place where it is called, and arranging for authentication to fail
7277 instead of killing the process with assert().
7279 75. The SPA client code was not working when the server requested OEM rather
7280 than Unicode encoding.
7282 76. Added code to make require_files with a specific uid setting more usable in
7283 the case where statting the file as root fails - usually a non-root-mounted
7284 NFS file system. When this happens and the failure is EACCES, Exim now
7285 forks a subprocess and does the per-uid checking as the relevant uid.
7287 77. Added process_log_path.
7289 78. If log_file_path was not explicitly set, a setting of check_log_space or
7290 check_log_inodes was ignored.
7292 79. If a space check for the spool or log partitions fails, the incident is now
7293 logged. Of course, in the latter case the data may get lost...
7295 80. Added the %p formatting code to string_format() so that it can be used to
7296 print addresses in debug_print(). Adjusted all the address printing in the
7297 debugging in store.c to use %p rather than %d.
7299 81. There was a concern that a line of code in smtp_in.c could overflow a
7300 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7301 initially expressed, the concern was not well-founded, because trailing
7302 spaces are removed early. However, if the trailing spaces were followed by
7303 a NULL, they did not get removed, so the overflow was possible. Two fixes
7306 (a) I re-wrote the offending code in a cleaner fashion.
7307 (b) If an incoming SMTP command contains a NULL character, it is rejected
7310 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7311 runs initgroups(), so that if the Exim user is in any additional groups,
7312 they will be used during message reception.
7318 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob