1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
10 HS/01 Handle trailing backslash gracefully. (CVE-2019-15846)
16 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
22 JH/01 Remove code calling the customisable local_scan function, unless a new
23 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
25 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
26 non-signal-safe functions being used.
28 JH/03 Bug 2269: When presented with a received message having a stupidly large
29 number of DKIM-Signature headers, disable DKIM verification to avoid
30 a resource-consumption attack. The limit is set at twenty.
32 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
33 report of oldest_pass in ${authres } in consequence, and separate out
34 some descriptions of reasons for verification fail.
36 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
37 files in the spool were present and unlocked. A queue-runner could spot
38 them, resulting in a duplicate delivery. Fix that by doing the unlock
39 after the unlink. Investigation by Tim Stewart. Take the opportunity to
40 add more error-checking on spoolfile handling while that code is being
43 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
44 No known attacks, no CVE, this is defensive hardening.
46 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
47 a queue-runner could start a delivery while other operations were ongoing.
48 Cutthrough delivery was a common victim, resulting in duplicate delivery.
49 Found and investigated by Tim Stewart. Fix by using the open message data
50 file handle rather than opening another, and not locally closing it (which
51 releases a lock) for that case, while creating the temporary .eml format
52 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
54 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
55 $sender_verify_failure/$recipient_verify_failure to "random".
57 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
60 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
61 Previously this would segfault.
63 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
66 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
67 like zero, since the resolver should be doing this for us, But we need one
68 as a CNAME but no MX presence gets the CNAME returned; we need to check
69 that doesn't point to an MX to declare it "no MX returned" rather than
70 "error, loop". A new main option is added so the older capability of
71 following some limited number of chain links is maintained.
73 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
75 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
76 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
77 now-deprecated earlier definitions used only specified the range up to TLS
78 1.2 (in the older-version library docs).
80 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
82 JH/15 Rework TLS client-side context management. Stop using a global, and
83 explicitly pass a context around. This enables future use of TLS for
84 connections to service-daemons (eg. malware scanning) while a client smtp
85 connection is using TLS; with cutthrough connections this is quite likely.
87 JH/16 Fix ARC verification to do AS checks in reverse order.
89 JH/17 Support a "tls" option on the ${readsocket } expansion item.
91 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
92 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
93 Previously the "utf8" would be re-prepended for every additional message.
95 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
96 Previously thery were accepted, resulting in issues when attempting to
97 forward messages to a non-supporting MTA.
99 PP/02 Let -n work with printing macros too, not just options.
101 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
102 one parent address was copied, and bogus data was used at delivery-logging
103 time. Either a crash (after delivery) or bogus log data could result.
104 Discovery and analysis by Tim Stewart.
106 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
107 Previously if the string ended mid-character, we did not insert the
108 promised '?' replacement.
110 PP/04 Documentation: current string operators work on bytes, not codepoints.
112 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
113 should pack well giving a smaller memory footprint so better caching and
114 therefore performance. Group the declarations where this can't be done so
115 that the byte-sized flag variables are not interspersed among pointer
116 variables, giving a better chance of good packing by the compiler.
118 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
119 non-null, to avoid issues with sites running BATV. Previously reports were
120 sent with an empty envelope sender so looked like bounces.
122 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
123 The ignore_error flag wasn't being returned from the filter subprocess so
124 was not set for later routers. Investigation and fix by Matthias Kurz.
126 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
127 and a msg:complete for the whole, when a message is manually removed using
128 -Mrm. Developement by Matthias Kurz, hacked on by JH.
130 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
131 a "Gnu special" function, asprintf() in the DB utility binary builds; I
132 hope that is portable enough.
134 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
135 requiring a known-CA anchor certificate; make it now rely entirely on the
136 TLSA as an anchor. Checking the name on the leaf cert against the name
137 on the A-record for the host is still done for TA (but not for EE mode).
139 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
140 included in delivery lines for non-proxied connections, when compiled with
141 SUPPORT_SOCKS and running with proxy logging enabled.
143 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
144 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
145 move the existing event to fire before the normal logging of message
146 failure so that custom logging is bracketed by normal logging.
148 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
149 msg:fail:internal event. Developement by Matthias Kurz.
151 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
152 far too small for todays use of crypto signatures stored there. Go all
153 the way to the max DNS message size of 64kB, even though this might be
154 overmuch for IOT constrained device use.
156 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
157 copy a string over itself. The library routine is documented as not
158 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
160 JH/32 For main options check_spool_space and check_inode_space, where the
161 platform supports 64b integers, support more than the previous 2^31 kB
162 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
163 the previous G, M, k.
165 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
166 $authenticated_fail_id variable on authentication failure. Previously
169 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
170 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
171 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
172 GNUTLS_SEC_PARAM_MEDIUM.
174 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
175 side. Previously we would continue as if no SNI had been received.
177 JH/36 Harden the handling of string-lists. When a list consisted of a sole
178 "<" character, which should be a list-separator specification, we walked
179 off past the nul-terimation.
181 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
182 causes) even when the retry time is not yet met. Previously they were
183 not, meaning that when (say) an account was over-quota and temp-rejecting,
184 and multiple senders' messages were queued, only one sender would get
185 notified on each configured delay_warning cycle.
187 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
189 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
190 cipher-suites, an error can be left on the stack even for a succeeding
191 accept; this results in impossible error messages when a later operation
194 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
195 return error codes indicating retry. Under TLS1.3 this becomes required.
197 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
198 it only wrote the new authenticators, resulting in a lack of tracking of
199 peer changes of ESMTP extensions until the next cache flush.
201 JH/41 Fix the loop reading a message header line to check for integer overflow,
202 and more-often against header_maxsize. Previously a crafted message could
203 induce a crash of the recive process; now the message is cleanly rejected.
205 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
206 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
212 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
213 When redis_servers is set to a list of > 1 element, and the Redis servers
214 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
215 case of MOVED into a DEFER case instead, thus moving the query onto the
216 next server in the list. For a cluster of N elements, all N servers must
217 be defined in redis_servers.
219 GF/02 Catch and remove uninitialized value warning in exiqsumm
220 Check for existence of @ARGV before looking at $ARGV[0]
222 JH/01 Replace the store_release() internal interface with store_newblock(),
223 which internalises the check required to safely use the old one, plus
224 the allocate and data copy operations duplicated in both (!) of the
225 extant use locations.
227 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
228 modifier. This matches the restriction on the commandline.
230 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
231 Previously only the last row was returned.
233 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
234 we assumed that tags in the header were well-formed, and parsed the
235 element content after inspecting only the first char of the tag.
236 Assumptions at that stage could crash the receive process on malformed
239 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
240 While running the DKIM ACL we operate on the Permanent memory pool so that
241 variables created with "set" persist to the DATA ACL. Also (at any time)
242 DNS lookups that fail create cache records using the Permanent pool. But
243 expansions release any allocations made on the current pool - so a dnsdb
244 lookup expansion done in the DKIM ACL releases the memory used for the
245 DNS negative-cache, and bad things result. Solution is to switch to the
246 Main pool for expansions.
247 While we're in that code, add checks on the DNS cache during store_reset,
248 active in the testsuite.
249 Problem spotted, and debugging aided, by Wolfgang Breyha.
251 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
252 When none of the hosts presented to a transport match an already-open
253 connection, close it and proceed with the list. Previously we would
254 queue the message. Spotted by Lena with Yahoo, probably involving
257 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
258 Previously a spurious "250 OK id=" response was appended to the proper
261 JH/08 The "support for" informational output now, which built with Content
262 Scanning support, has a line for the malware scanner interfaces compiled
263 in. Interface can be individually included or not at build time.
265 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
266 by the template makefile "src/EDITME". The "STREAM" support for an older
267 ClamAV interface method is removed.
269 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
270 rows affected is given instead).
272 JH/11 The runtime Berkeley DB library version is now additionally output by
273 "exim -d -bV". Previously only the compile-time version was shown.
275 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
276 SMTP connection. Previously, when one had more receipients than the
277 first, an abortive onward connection was made. Move to full support for
278 multiple onward connections in sequence, handling cutthrough connection
279 for all multi-message initiating connections.
281 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
282 routers. Previously, a multi-recipient message would fail to match the
283 onward-connection opened for the first recipient, and cause its closure.
285 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
286 a timeout on read on a GnuTLS initiating connection, resulting in the
287 initiating connection being dropped. This mattered most when the callout
288 was marked defer_ok. Fix to keep the two timeout-detection methods
291 JH/15 Relax results from ACL control request to enable cutthrough, in
292 unsupported situations, from error to silently (except under debug)
293 ignoring. This covers use with PRDR, frozen messages, queue-only and
296 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
298 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
299 metadata, resulting in a crash in free().
301 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
302 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
303 Broken also in d185889f4, with init system revamp.
305 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
306 Previously we abruptly closed the connection after reading a malware-
307 found indication; now we go on to read the "scan ok" response line,
310 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
311 ACL. Previously, a crash would result.
313 JH/19 Speed up macro lookups during configuration file read, by skipping non-
314 macro text after a replacement (previously it was only once per line) and
315 by skipping builtin macros when searching for an uppercase lead character.
317 JH/20 DANE support moved from Experimental to mainline. The Makefile control
318 for the build is renamed.
320 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
321 was allocated for every new TLS startup, meaning one per message. Fix
322 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
324 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
325 reported the original. Fix to report (as far as possible) the ACL
326 result replacing the original.
328 JH/23 Fix memory leak during multi-message connections using STARTTLS under
329 OpenSSL. Certificate information is loaded for every new TLS startup,
330 and the resources needed to be freed.
332 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
334 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
337 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
338 DATA response info to the (existing) per-recipient response info for
339 the "C=" log element. It can have useful tracking info from the
340 destination system. Patch from Simon Arlott.
342 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
343 length value. Previously this would segfault.
345 HS/02 Support Avast multiline protoocol, this allows passing flags to
346 newer versions of the scanner.
348 JH/28 Ensure that variables possibly set during message acceptance are marked
349 dead before release of memory in the daemon loop. This stops complaints
350 about them when the debug_store option is enabled. Discovered specifically
351 for sender_rate_period, but applies to a whole set of variables.
352 Do the same for the queue-runner and queue-list loops, for variables set
353 from spool message files. Do the same for the SMTP per-message loop, for
354 certain variables indirectly set in ACL operations.
356 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
357 as a multi-recipient message from a mailinglist manager). The coding had
358 an arbitrary cutoff number of characters while checking for more input;
359 enforced by writing a NUL into the buffer. This corrupted long / fast
360 input. The problem was exposed more widely when more pipelineing of SMTP
361 responses was introduced, and one Exim system was feeding another.
362 The symptom is log complaints of SMTP syntax error (NUL chars) on the
363 receiving system, and refused recipients seen by the sending system
364 (propating to people being dropped from mailing lists).
365 Discovered and pinpointed by David Carter.
367 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
368 replaced by the ${authresults } expansion.
370 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
372 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
373 allows proper process termination in container environments.
375 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
376 Previously the "final dot" had a newline after it; ensure it is CR,LF.
378 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
379 and "err_perm", deprecated since 4.83 when the RFC-defined words
380 "temperror" and "permerror" were introduced.
382 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
383 transport-filters or DKIM-signing. The restriction was lost in the
384 consolidation of verify-callout and delivery SMTP handling.
385 Extend the restriction to also cover ARC-signing.
387 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
388 in defer=pass mode supply a 450 to the initiator. Previously the message
391 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
392 tls_require_ciphers is used as before.
394 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
395 "pass_unscanned". Only tmpfails from the scanner are written to
396 the paniclog, as they may require admin intervention (permission
397 denied, license issues). Other scanner errors (like decompression
398 bombs) do not cause a paniclog entry.
400 JH/36 Fix reinitialisation of DKIM logging variable between messages.
401 Previously it was possible to log spurious information in receive log
404 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
405 triggered odd behaviour from Outlook Express clients.
407 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
410 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
411 since the IETF WG has not yet settled on that versus the original
412 "bare" representation.
414 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
415 Previously the millisecond value corrupted the output.
416 Fix also for syslog_pid=no and log_selector +pid, for which the pid
417 corrupted the output.
423 JH/01 Rework error string handling in TLS interface so that the caller in
424 more cases is responsible for logging. This permits library-sourced
425 string to be attached to addresses during delivery, and collapses
426 pairs of long lines into single ones.
428 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
429 during configuration. Wildcards are allowed and expanded.
431 JH/02 Rework error string handling in DKIM to pass more info back to callers.
432 This permits better logging.
434 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
435 do not close it down and have the child transport start it up again on
436 the passed-on TCP connection. Instead, proxy the child (and any
437 subsequent ones) for TLS via a unix-domain socket channel. Logging is
438 affected: the continued delivery log lines do not have any DNSSEC, TLS
439 Certificate or OCSP information. TLS cipher information is still logged.
441 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
442 identical IP addresses on different listening ports. Will also affect
445 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
446 add noisy ifdef guards to special-case this sillyness.
447 Patch from Bernd Kuhls.
449 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
450 than 255 are no longer allowed.
452 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
453 Disable the session-cache too, which might reduce our load. Since we
454 currrectly use a new context for every connection, both as server and
455 client, there is no benefit for these.
456 GnuTLS appears to not support tickets server-side by default (we don't
457 call gnutls_session_ticket_enable_server()) but client side is enabled
458 by default on recent versions (3.1.3 +) unless the PFS priority string
461 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
462 <https://reproducible-builds.org/specs/source-date-epoch/>.
464 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
465 the check for any unsuccessful recipients did not notice the limit, and
466 erroneously found still-pending ones.
468 JH/08 Pipeline CHUNKING command and data together, on kernels that support
469 MSG_MORE. Only in-clear (not on TLS connections).
471 JH/09 Avoid using a temporary file during transport using dkim. Unless a
472 transport-filter is involved we can buffer the headers in memory for
473 creating the signature, and read the spool data file once for the
474 signature and again for transmission.
476 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
477 4.77 as the kernel support then wasn't solid, having issues in 64bit
478 mode. Now, it's been long enough. Add support for FreeBSD also.
480 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
481 case where the routing stage had gathered several addresses to send to
482 a host before calling the transport for the first, we previously failed
483 to close down TLS in the old transport process before passing the TCP
484 connection to the new process. The new one sent a STARTTLS command
485 which naturally failed, giving a failed delivery and bloating the retry
486 database. Investigation and fix prototype from Wolfgang Breyha.
488 JH/12 Fix check on SMTP command input synchronisation. Previously there were
489 false-negatives in the check that the sender had not preempted a response
490 or prompt from Exim (running as a server), due to that code's lack of
491 awareness of the SMTP input buffering.
493 PP/04 Add commandline_checks_require_admin option.
494 Exim drops privileges sanely, various checks such as -be aren't a
495 security problem, as long as you trust local users with access to their
496 own account. When invoked by services which pass untrusted data to
497 Exim, this might be an issue. Set this option in main configuration
498 AND make fixes to the calling application, such as using `--` to stop
501 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
502 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
503 responses to those, into a single TLS record each way (this usually means
504 a single packet). As a side issue, smtp_enforce_sync now works on TLS
507 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
508 affects you only if you're dancing at the edge of the param size limits.
509 If you are, and this message makes sense to you, then: raise the
510 configured limit or use OpenSSL 1.1. Nothing we can do for older
513 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
514 cmdline element to get the documented default one. Previously it was
517 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
518 are unsuable for non-message-specific reasons - eg. connection timeout,
519 banner-time rejection.
521 JH/16 Fix logging of delivery remote port, when specified by router, under
524 PP/06 Repair manualroute's ability to take options in any order, even if one
525 is the name of a transport.
528 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
530 JH/17 Change the list-building routines interface to use the expanding-string
531 triplet model, for better allocation and copying behaviour.
533 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
534 Previously it was constructed the first time a possibly-matching string
535 was met in the configuration file input during startup; now it is done
538 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
539 compatible one, to avoid the (poorly documented) possibility of a config
540 file in the working directory redirecting the DB files, possibly correpting
541 some existing file. CVE-2017-10140 assigned for BDB.
543 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
544 cache-hot. Previously, although the result was properly cached, the
545 initial verify call returned a defer.
547 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
548 the main verify for receipient in uncached-mode.
550 JH/22 Retire historical build files to an "unsupported" subdir. These are
551 defined as "ones for which we have no current evidence of testing".
553 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
554 if present. Previously it was ignored.
556 JH/24 Start using specified-initialisers in C structure init coding. This is
557 a C99 feature (it's 2017, so now considered safe).
559 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
560 if was a fixed-sized field and bitmask ops via macros; it is now more
563 PP/07 GitHub PR 56: Apply MariaDB build fix.
564 Patch provided by Jaroslav Škarvada.
566 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
567 during Coverity cleanups [4.87 JH/47]
568 Diagnosis and fix provided by Michael Fischer v. Mollard.
570 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
571 the right size to place the terminating semicolon on its own folded
572 line, the header hash was calculated to an incorrect value thanks to
573 the (relaxed) space the fold became.
575 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
576 and confused the parent.
578 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
579 which could crash as a result. This could lead to undeliverable messages.
581 JH/28 Logging: "next input sent too soon" now shows where input was truncated
584 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
585 matters on fast-turnover and PID-randomising systems, which were getting
586 out-of-order delivery.
588 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
589 a possibly-overlapping copy. The symptom was that "Remote host closed
590 connection in response to HELO" was logged instead of the actual 4xx
593 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
594 Previously only that bufferd was discarded, resulting in SYMTP command
597 JH/32 DKIM: when a message has multiple signatures matching an identity given
598 in dkim_verify_signers, run the dkim acl once for each. Previously only
599 one run was done. Bug 2189.
601 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
602 "panic the current process" to "deliberately defer". The panic log is
603 still written with the problem list name; the mail and reject logs now
604 get a temp-reject line for the message that was being handled, saying
605 something like "domains check lookup or other defer". The SMTP 451
606 message is still "Temporary local problem".
608 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
609 A crafted sequence of BDAT commands could result in in-use memory beeing
610 freed. CVE-2017-16943.
612 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
613 from SMTP input. Previously it was always done; now only done for DATA
614 and not BDAT commands. CVE-2017-16944.
616 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
617 to the message (such as an overlong header line). Previously this was
618 not done and we did not exit BDAT mode. Followon from the previous item
619 though a different problem.
625 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
626 than -2003 did; needs libidn2 in addition to libidn.
628 JH/02 The path option on a pipe transport is now expanded before use.
630 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
631 Patch provided by "Björn", documentation fix added too.
633 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
634 missing a wire-to-host endian conversion.
636 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
637 close after a BDAT command line could be taken as a following command,
638 giving a synch failure. Fix by only checking for synch immediately
639 before acknowledging the chunk.
641 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
642 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
643 macro. Patches provided by Josh Soref.
645 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
646 Previously we did not; the RFC seems ambiguous and VRFY is not listed
647 by IANA as a service extension. However, John Klensin suggests that we
650 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
651 the dkim code may be unix-mode line endings rather than smtp wire-format
652 CRLF, so prepend a CR to any bare LF.
654 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
655 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
657 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
658 the first were themselves being wrongly included in the feed into dkim
659 processing; with most chunk sizes in use this resulted in an incorrect
660 body hash calculated value.
662 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
663 DKIM signature block, for verification. Although advised against by
664 standards it is specifically not ruled illegal.
666 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
668 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
669 missing a body hash (the bh= tag).
671 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
672 It seems that HAProxy sends the Proxy Protocol information in clear and
673 only then does a TLS startup, so do the same.
675 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
676 TCP connections (such as for Spamd) unless the daemon successfully set
677 Fast Open mode on its listening sockets. This fixes breakage seen on
678 too-old kernels or those not configured for Fast Open, at the cost of
679 requiring both directions being enabled for TFO, and TFO never being used
680 by non-daemon-related Exim processes.
682 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
683 endings, at least on the first header line. Try to canonify any that get
684 past that check, despite the cost.
686 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
687 now limited to an arbitrary five deep, while parsing addresses with the
688 strip_excess_angle_brackets option enabled.
690 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
691 instead leave the unprompted TLS handshake in socket buffer for the
692 TLS library to consume.
694 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
696 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
698 JH/16 Drop variables when they go out of scope. Memory management drops a whole
699 region in one operation, for speed, and this leaves assigned pointers
700 dangling. Add checks run only under the testsuite which checks all
701 variables at a store-reset and panics on a dangling pointer; add code
702 explicitly nulling out all the variables discovered. Fixes one known
703 bug: a transport crash, where a dangling pointer for $sending_ip_address
704 originally assigned in a verify callout, is re-used.
706 PP/06 Drop '.' from @INC in various Perl scripts.
708 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
710 PP/08 Reduce a number of compilation warnings under clang; building with
711 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
712 should be warning-free.
714 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
716 HS/01 Fix portability problems introduced by PP/08 for platforms where
717 realloc(NULL) is not equivalent to malloc() [SunOS et al].
719 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
720 chunk. This allows us to accept broken chunked messages. We need a more
721 general solution here.
723 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
724 already-broken messages in the queue.
726 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
728 JH/19 Fix reference counting bug in routing-generated-address tracking.
734 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
735 supports it and a size is available (ie. the sending peer gave us one).
737 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
738 years of being deprecated). The replacements are the ACLs
739 acl_smtp_mime and acl_not_smtp_mime.
741 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
742 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
743 or even an in-clear connection were permitted. Now, if the host lookup
744 was dnssec and dane was requested then the host is only used if the
745 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
746 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
747 if one fails this test.
748 This means that a poorly-configured remote DNS will make it incommunicado;
749 but it protects against a DNS-interception attack on it.
751 JH/04 Bug 1810: make continued-use of an open smtp transport connection
752 non-noisy when a race steals the message being considered.
754 JH/05 If main configuration option tls_certificate is unset, generate a
755 self-signed certificate for inbound TLS connections.
757 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
758 in rewrites and routers.
760 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
761 and logged a warning sing 4.83; now they are a configuration file error.
763 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
764 (lacking @domain). Apply the same qualification processing as RCPT.
766 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
768 JH/10 Support ${sha256:} applied to a string (as well as the previous
771 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
772 a cutthrough deliver is pending, as we always want to make a connection.
773 This also avoids re-routing the message when later placing the cutthrough
774 connection after a verify cache hit.
775 Do not update it with the verify result either.
777 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
778 when routing results in more than one destination address.
780 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
781 signing (which inhibits the cutthrough capability). Previously only
782 the presence of an option was tested; now an expansion evaluating as
783 empty is permissible (obviously it should depend only on data available
784 when the cutthrough connection is made).
786 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
787 the relevant preceding SMTP command did not note the pipelining mode.
789 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
790 Previously they were not counted.
792 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
793 as one having no matching records. Previously we deferred the message
794 that needed the lookup.
796 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
797 distinguished as "(=".
799 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
800 for missing MX records. Previously it only worked for missing A records.
802 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
804 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
805 after the data-go-ahead and data-ack. Patch from Jason Betts.
807 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
808 even for a "none" policy. Patch from Tony Meyer.
810 JH/22 Fix continued use of a connection for further deliveries. If a port was
811 specified by a router, it must also match for the delivery to be
814 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
815 When one of the recipients of a message was unsuitable for the connection
816 (has no matching addresses), we lost track of needing to mark it
817 deferred. As a result mail would be lost.
819 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
821 JH/25 Decoding ACL controls is now done using a binary search; the source code
822 takes up less space and should be simpler to maintain. Merge the ACL
823 condition decode tables also, with similar effect.
825 JH/26 Fix problem with one_time used on a redirect router which returned the
826 parent address unchanged. A retry would see the parent address marked as
827 delivered, so not attempt the (identical) child. As a result mail would
830 JH/27 Fix a possible security hole, wherein a process operating with the Exim
831 UID can gain a root shell. Credit to http://www.halfdog.net/ for
832 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
835 JH/28 Enable {spool,log} filesystem space and inode checks as default.
836 Main config options check_{log,spool}_{inodes,space} are now
837 100 inodes, 10MB unless set otherwise in the configuration.
839 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
840 Previously it only applied to the main-section connection policy
843 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
845 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
846 by me. Added RFC7919 DH primes as an alternative.
848 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
849 are not in the system include path.
851 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
852 GnuTLS, when a session startup failed (eg because the client disconnected)
853 Exim did stdio operations after fclose. This was exposed by a recent
854 change which nulled out the file handle after the fclose.
856 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
857 signed directly by the cert-signing cert, rather than an intermediate
858 OCSP-signing cert. This is the model used by LetsEncrypt.
860 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
862 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
863 an incoming connection.
865 HS/02 Bug 1802: Do not half-close the connection after sending a request
868 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
869 fallback to "prime256v1".
871 JH/34 SECURITY: Use proper copy of DATA command in error message.
872 Could leak key material. Remotely exploitable. CVE-2016-9963.
878 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
879 and 3.4.4 - once the server is enabled to respond to an OCSP request
880 it does even when not requested, resulting in a stapling non-aware
881 client dropping the TLS connection.
883 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
884 support variable-length bit vectors. No functional change.
886 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
887 The I= interface field on outgoing lines is now after the H= remote
888 host field, same as incoming lines. There is a separate
889 outgoing_interface log selector which allows you to disable the
892 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
893 If not running log_selector +smtp_connection the mainlog would be held
894 open indefinitely after a "too many connections" event, including to a
895 deleted file after a log rotate. Leave the per net connection logging
896 leaving it open for efficiency as that will be quickly detected by the
897 check on the next write.
899 HS/01 Bug 1671: Fix post transport crash.
900 Processing the wait-<transport> messages could crash the delivery
901 process if the message IDs didn't exist for some reason. When
902 using 'split_spool_directory=yes' the construction of the spool
903 file name failed already, exposing the same netto behaviour.
905 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
906 mime_regex ACL conditions.
908 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
909 to DSN fail messages (bounces): remote IP, remote greeting, remote response
910 to HELO, local diagnostic string.
912 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
913 log line to debug. Even when configured with a tls authenticator many
914 client connections are expected to not authenticate in this way, so
915 an authenticate fail is not an error.
917 HS/02 Add the Exim version string to the process info. This way exiwhat
918 gives some more detail about the running daemon.
920 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
921 matter for fast-change records such as DNSBLs.
923 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
924 retry record, even if constant. There may be multiple transports with
925 different interface settings and the retry behaviour needs to be kept
928 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
930 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
932 JH/10 Bug 840: fix log_defer_output option of pipe transport
934 JH/11 Bug 830: use same host for all RCPTS of a message, even under
935 hosts_randomize. This matters a lot when combined with mua_wrapper.
937 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
938 ${quote_pgsql:<string>} operator.
940 JH/13 Bug 1708: avoid misaligned access in cached lookup.
942 JH/14 Change header file name for freeradius-client. Relevant if compiling
943 with Radius support; from the Gentoo tree and checked under Fedora.
945 JH/15 Bug 1712: Introduce $prdr_requested flag variable
947 JH/16 Bug 1714: Permit an empty string as expansion result for transport
948 option transport_filter, meaning no filtering.
950 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
952 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
953 defaults to "*" (all hosts). The variable is now available when not built
954 with TLS, default unset, mainly to enable keeping the testsuite sane.
955 If a server certificate is not supplied (via tls_certificate) an error is
956 logged, and clients will find TLS connections fail on startup. Presumably
957 they will retry in-clear.
958 Packagers of Exim are strongly encouraged to create a server certificate
959 at installation time.
961 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
962 with the $config_file variable.
964 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
965 in transport context, after the attempt, and per-recipient. The latter type
966 is per host attempted. The event data is the error message, and the errno
967 information encodes the lookup type (A vs. MX) used for the (first) host,
968 and the trailing two digits of the smtp 4xx response.
970 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
971 to write to mainlog (or rejectlog, paniclog) in the window between file
972 creation and permissions/ownership being changed. Particularly affects
973 installations where exicyclog is run as root, rather than exim user;
974 result is that the running daemon panics and dies.
976 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
978 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
979 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
980 "pri" and "weight". Note that the previous implicit priority given by the
981 list order is no longer honoured.
983 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
986 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
987 by defining SUPPORT_SOCKS.
989 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
990 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
991 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
992 variables are renamed to proxy_{local,external}_{address,port}.
994 JH/27 Move Internationalisation support from Experimental to mainline, enabled
995 for a build by defining SUPPORT_I18N
997 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
998 of the query string, and make ${quote_redis:} do that quoting.
1000 JH/29 Move Events support from Experimental to mainline, enabled by default
1001 and removable for a build by defining DISABLE_EVENT.
1003 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1005 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1006 cached by the daemon.
1008 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1009 by defining LOOKUP_REDIS. The libhiredis library is required.
1011 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1012 keys are given for lookup.
1014 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1015 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1016 only supported when built with TLS support. The PolarSSL SHA routines
1017 are still used when the TLS library is too old for convenient support.
1019 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1020 openssl_options), for security. OpenSSL forces this from version 1.1.0
1021 server-side so match that on older versions.
1023 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1024 allocation for $value could be released as the expansion processing
1025 concluded, but leaving the global pointer active for it.
1027 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1028 and to use the domains and local_parts ACL conditions.
1030 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1031 incorrectly not doubled on cutthrough transmission, hence seen as a
1032 body-termination at the receiving system - resulting in truncated mails.
1033 Commonly the sender saw a TCP-level error, and retransmitted the message
1034 via the normal store-and-forward channel. This could result in duplicates
1035 received - but deduplicating mailstores were liable to retain only the
1036 initial truncated version.
1038 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1040 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1042 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1043 we're in there, support oversigning also; bug 1309.
1045 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1047 HS/04 Add support for keep_environment and add_environment options.
1049 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1050 either intentional arithmetic overflow during PRNG, or testing config-
1053 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1054 delivery resulted in actual delivery. Cancel cutthrough before DATA
1057 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1058 rejects a recipient: pass the reject to the originator.
1060 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1061 Many were false-positives and ignorable, but it's worth fixing the
1064 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1065 for the new environment-manipulation done at startup. Move the routines
1066 from being local to tls.c to being global via the os.c file.
1068 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1069 an extract embedded as result-arg for a map, the first arg for extract
1070 is unavailable so we cannot tell if this is a numbered or keyed
1071 extraction. Accept either.
1077 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1080 JH/02 The smtp transport option "multi_domain" is now expanded.
1082 JH/03 The smtp transport now requests PRDR by default, if the server offers
1085 JH/04 Certificate name checking on server certificates, when exim is a client,
1086 is now done by default. The transport option tls_verify_cert_hostnames
1087 can be used to disable this per-host. The build option
1088 EXPERIMENTAL_CERTNAMES is withdrawn.
1090 JH/05 The value of the tls_verify_certificates smtp transport and main options
1091 default to the word "system" to access the system default CA bundle.
1092 For GnuTLS, only version 3.0.20 or later.
1094 JH/06 Verification of the server certificate for a TLS connection is now tried
1095 (but not required) by default. The verification status is now logged by
1096 default, for both outbound TLS and client-certificate supplying inbound
1099 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1102 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1103 Status Notification (bounce) messages are now MIME format per RFC 3464.
1104 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1105 under the control of the dsn_advertise_hosts option, and routers may
1106 have a dsn_lasthop option.
1108 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1109 default, modifiable by a malware= option. The list separator for
1110 the options can now be changed in the usual way. Bug 68.
1112 JH/10 The smtp_receive_timeout main option is now expanded before use.
1114 JH/11 The incoming_interface log option now also enables logging of the
1115 local interface on delivery outgoing connections.
1117 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1118 if the interface and destination host and port all match.
1120 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1123 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1124 Patch from Andrew Lewis.
1126 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1127 now supports optional time-restrictions, weighting, and priority
1128 modifiers per server. Patch originally by <rommer@active.by>.
1130 JH/16 The spamd_address main option now supports a mixed list of local
1131 and remote servers. Remote servers can be IPv6 addresses, and
1132 specify a port-range.
1134 JH/17 Bug 68: The spamd_address main option now supports an optional
1135 timeout value per server.
1137 JH/18 Bug 1581: Router and transport options headers_add/remove can
1138 now have the list separator specified.
1140 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1143 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1146 JH/21 Support for the A6 type of dns record is withdrawn.
1148 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1149 rather than the verbs used.
1151 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1152 from 255 to 1024 chars.
1154 JH/24 Verification callouts now attempt to use TLS by default.
1156 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1157 are generic router options now. The defaults didn't change.
1159 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1160 Original patch from Alexander Shikoff, worked over by JH.
1162 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1163 files if ZCAT_COMMAND is not executable.
1165 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1167 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1169 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1170 Normally benign, it bites when the pair was led to by a CNAME;
1171 modern usage is to not canonicalize the domain to a CNAME target
1172 (and we were inconsistent anyway for A-only vs AAAA+A).
1174 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1176 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1177 when evaluating $sender_host_dnssec.
1179 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1180 $sender_helo_dnssec variable.
1182 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1184 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1186 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1188 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1189 documented as working, but never had. Support all but $spam_report.
1191 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1192 added for tls authenticator.
1194 HS/03 Add perl_taintmode main config option
1200 TL/01 When running the test suite, the README says that variables such as
1201 no_msglog_check are global and can be placed anywhere in a specific
1202 test's script, however it was observed that placement needed to be near
1203 the beginning for it to behave that way. Changed the runtest perl
1204 script to read through the entire script once to detect and set these
1205 variables, reset to the beginning of the script, and then run through
1206 the script parsing/test process like normal.
1208 TL/02 The BSD's have an arc4random API. One of the functions to induce
1209 adding randomness was arc4random_stir(), but it has been removed in
1210 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1211 function when detected.
1213 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1214 cause callback expansion.
1216 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1217 syntax errors in an expansion can be treated as a string instead of
1218 logging or causing an error, due to the internal use of bool_lax
1219 instead of bool when processing it.
1221 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1222 server certificates when making smtp deliveries.
1224 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1226 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1228 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1230 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1231 Merged patch from Sebastian Wiedenroth.
1233 JH/05 Fix results-pipe from transport process. Several recipients, combined
1234 with certificate use, exposed issues where response data items split
1235 over buffer boundaries were not parsed properly. This eventually
1236 resulted in duplicates being sent. This issue only became common enough
1237 to notice due to the introduction of connection certificate information,
1238 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1240 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1241 size buffer was used, resulting in syntax errors when an expansion
1244 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1245 version 3.3.6 or later.
1247 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1248 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1249 both become "event_action", the variables become $event_name, $event_data
1250 and $event_defer_errno. There is a new variable $verify_mode, usable in
1251 routers, transports and related events. The tls:cert event is now also
1252 raised for inbound connections, if the main configuration event_action
1255 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1256 early OCSP support, but no stapling (appears to be less than 1.0.0).
1258 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1259 server certificate names available under the smtp transport option
1260 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1263 JH/10 Time-related extraction expansions from certificates now use the main
1264 option "timezone" setting for output formatting, and are consistent
1265 between OpenSSL and GnuTLS compilations. Bug 1541.
1267 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1268 encoded parameter in the incoming message. Bug 1558.
1270 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1271 include certificate info, eximon was claiming there were spoolfile
1274 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1276 JH/14 Log delivery-related information more consistently, using the sequence
1277 "H=<name> [<ip>]" wherever possible.
1279 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1280 are problematic for Debian distribution, omit them from the release
1283 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1285 JH/16 Fix string representation of time values on 64bit time_t architectures.
1288 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1289 output list separator was used.
1294 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1295 checkers that were complaining about end of non-void function with no
1298 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1299 This was a regression introduced in 4.83 by another bugfix.
1301 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1303 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1304 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1310 TF/01 Correctly close the server side of TLS when forking for delivery.
1312 When a message was received over SMTP with TLS, Exim failed to clear up
1313 the incoming connection properly after forking off the child process to
1314 deliver the message. In some situations the subsequent outgoing
1315 delivery connection happened to have the same fd number as the incoming
1316 connection previously had. Exim would try to use TLS and fail, logging
1317 a "Bad file descriptor" error.
1319 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1320 utilities have not been installed.
1322 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1323 temporary space as the ACL may create new global variables.
1325 TL/01 LDAP support uses per connection or global context settings, depending
1326 upon the detected version of the libraries at build time.
1328 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1329 to extract and use the src ip:port in logging and expansions as if it
1330 were a direct connection from the outside internet. PPv2 support was
1331 updated based on HAProxy spec change in May 2014.
1333 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1335 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1336 Properly escape header and check for NULL return.
1338 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1341 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1343 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1344 characters in header names, implemented as a verify condition.
1345 Contributed by Michael Fischer v. Mollard.
1347 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1348 results permerror and temperror. Previous values are deprecated but
1349 still accepted. In a future release, err_perm and err_temp will be
1350 completely removed, which will be a backward incompatibility if the
1351 ACL tests for either of these two old results. Patch contributed by
1352 user bes-internal on the mailing list.
1354 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1356 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1357 selectors, in both main and reject logs.
1359 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1362 JH/07 Add malware type "sock" for talking to simple daemon.
1364 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1366 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1367 routers/transports under cutthrough routing.
1369 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1370 numbers. Touch up "bool" conditional to keep the same definition.
1372 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1374 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1376 JH/12 Expand items in router/transport headers_add or headers_remove lists
1377 individually rather than the list as a whole. Bug 1452.
1379 Required for reasonable handling of multiple headers_ options when
1380 they may be empty; requires that headers_remove items with embedded
1381 colons must have them doubled (or the list-separator changed).
1383 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1384 view the policy declared in the DMARC record. Currently, $dmarc_status
1385 is a combined value of both the record presence and the result of the
1388 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1390 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1391 dnslookup router and the smtp transport (applying to the forward
1394 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1395 of ldap servers used for a specific lookup. Patch provided by Heiko
1398 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1399 New variable $lookup_dnssec_authenticated for observability.
1401 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1402 Patch submitted by Lars Timman.
1404 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1406 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1407 Requires trusted mode and valid format message id, aborts otherwise.
1408 Patch contributed by Heiko Schlichting.
1410 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1411 certextract with support for various fields. Bug 1358.
1413 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1414 is requested by default, modifiable by smtp transport option
1417 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1418 operate on certificate variables to give certificate fingerprints
1419 Also new ${sha256:cert_variable}.
1421 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1423 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1424 Christian Aistleitner.
1426 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1428 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1429 file. Patch from Wolfgang Breyha.
1431 JH/25 Expand the coverage of the delivery $host and $host_address to
1432 client authenticators run in verify callout. Bug 1476.
1434 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1435 align with daemon_smtp_ports. Bug 72.
1437 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1438 support and error reporting did not work properly.
1440 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1441 and is readable. Patch from Andrew Colin Kissa.
1443 TL/14 Enhance documentation of ${run expansion and how it parses the
1444 commandline after expansion, particularly in the case when an
1445 unquoted variable expansion results in an empty value.
1447 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1449 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1450 by renaming away from it. Addresses GH issue 15, reported by
1453 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1455 TL/15 SECURITY: prevent double expansion in math comparison functions
1456 (can expand unsanitized data). Not remotely exploitable.
1463 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1465 PP/02 Make -n do something, by making it not do something.
1466 When combined with -bP, the name of an option is not output.
1468 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1471 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1472 $sender_host_name and config options to manage this, and basic check
1475 PP/05 DSCP support for outbound connections and control modifier for inbound.
1477 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1478 (Only plugin which currently uses this is kerberos4, which nobody should
1479 be using, but we should make it available and other future plugins might
1480 conceivably use it, even though it would break NAT; stuff *should* be
1481 using channel bindings instead).
1483 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1484 name; added for Sendmail compatibility; requires admin caller.
1485 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1486 just ignore it); requires trusted caller.
1487 Also parse but ignore: -Ac -Am -X<logfile>
1490 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1492 TL/02 Add +smtp_confirmation as a default logging option.
1494 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1495 Patch by Magnus Holmgren from 2007-02-20.
1497 TL/04 Bugzilla 1281 - Spec typo.
1498 Bugzilla 1283 - Spec typo.
1499 Bugzilla 1290 - Spec grammar fixes.
1501 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1503 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1505 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1506 mailing list by by Dmitry Isaikin.
1508 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1510 JH/02 Support "G" suffix to numbers in ${if comparisons.
1512 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1514 NM/01 Bugzilla 1197 - Spec typo
1515 Bugzilla 1196 - Spec examples corrections
1517 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1519 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1520 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1523 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1524 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1526 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1527 "acl {{name}{arg}...}", and optional args on acl condition
1530 JH/05 Permit multiple router/transport headers_add/remove lines.
1532 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1534 JH/07 Avoid using a waiting database for a single-message-only transport.
1535 Performance patch from Paul Fisher. Bugzilla 1262.
1537 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1540 JH/09 Add $headers_added variable, with content from use of ACL modifier
1541 add_header (but not yet added to the message). Bugzilla 199.
1543 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1544 Pulled from Bugzilla 817 by Wolfgang Breyha.
1546 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1548 (nb: this is the same fix as in Exim 4.80.1)
1550 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1553 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1556 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1558 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1559 advertise SMTP AUTH mechanism to us, instead of a generic
1560 protocol violation error. Also, make Exim more robust to bad
1561 data from the Dovecot auth socket.
1563 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1565 When a queue runner is handling a message, Exim first routes the
1566 recipient addresses, during which it prunes them based on the retry
1567 hints database. After that it attempts to deliver the message to
1568 any remaining recipients. It then updates the hints database using
1571 So if a recipient address works intermittently, it can get repeatedly
1572 deferred at routing time. The retry hints record remains fresh so the
1573 address never reaches the final cutoff time.
1575 This is a fairly common occurrence when a user is bumping up against
1576 their storage quota. Exim had some logic in its local delivery code
1577 to deal with this. However it did not apply to per-recipient defers
1578 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1580 This change adds a proper retry rule check during routing so that the
1581 final cutoff time is checked against the message's age. We only do
1582 this check if there is an address retry record and there is not a
1583 domain retry record; this implies that previous attempts to handle
1584 the address had the retry_use_local_parts option turned on. We use
1585 this as an approximation for the destination being like a local
1586 delivery, as in LMTP.
1588 I suspect this new check makes the old local delivery cutoff check
1589 redundant, but I have not verified this so I left the code in place.
1591 TF/02 Correct gecos expansion when From: is a prefix of the username.
1593 Test 0254 submits a message to Exim with the header
1597 When I ran the test suite under the user fanf2, Exim expanded
1598 the header to contain my full name, whereas it should have added
1599 a Resent-Sender: header. It erroneously treats any prefix of the
1600 username as equal to the username.
1602 This change corrects that bug.
1604 GF/01 DCC debug and logging tidyup
1605 Error conditions log to paniclog rather than rejectlog.
1606 Debug lines prefixed by "DCC: " to remove any ambiguity.
1608 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1610 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1611 Bug spotted by Jeremy Harris; was flawed since initial commit.
1612 Would have resulted in OCSP responses post-SNI triggering an Exim
1613 NULL dereference and crash.
1615 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1617 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1618 Bug detection, analysis and fix by Samuel Thibault.
1619 Bugzilla 1331, Debian bug #698092.
1621 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1623 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1624 Server implementation by Todd Lyons, client by JH.
1625 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1626 config variable "prdr_enable" controls whether the server
1627 advertises the facility. If the client requests PRDR a new
1628 acl_data_smtp_prdr ACL is called once for each recipient, after
1629 the body content is received and before the acl_smtp_data ACL.
1630 The client is controlled by both of: a hosts_try_prdr option
1631 on the smtp transport, and the server advertisement.
1632 Default client logging of deliveries and rejections involving
1633 PRDR are flagged with the string "PRDR".
1635 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1636 fclose(). Diagnosis by Todd Lyons.
1638 PP/17 Update configure.default to handle IPv6 localhost better.
1639 Patch by Alain Williams (plus minor tweaks).
1642 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1643 This is now consistent with GnuTLS, and is now documented: the
1644 previous undocumented portable approach to treating the option as
1645 unset was to force an expansion failure. That still works, and
1646 an empty string is now equivalent.
1648 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1649 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1650 not performing validation itself.
1652 PP/20 Added force_command boolean option to pipe transport.
1653 Patch from Nick Koston, of cPanel Inc.
1655 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1658 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1660 PP/21 Fix eximon continuous updating with timestamped log-files.
1661 Broken in a format-string cleanup in 4.80, missed when I repaired the
1662 other false fix of the same issue.
1663 Report and fix from Heiko Schlichting.
1666 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1667 Report from Prashanth Katuri.
1669 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1670 It's SecureTransport, so affects any MacOS clients which use the
1671 system-integrated TLS libraries, including email clients.
1673 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1674 using a MIME ACL for non-SMTP local injection.
1675 Report and assistance in diagnosis by Warren Baker.
1677 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1679 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1681 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1682 last id that failed so it may be referenced in subsequent ACL's.
1684 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1687 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1688 ldap library initialization, allowing self-signed CA's to be
1689 used. Also properly sets require_cert option later in code by
1690 using NULL (global ldap config) instead of ldap handle (per
1691 session). Bug diagnosis and testing by alxgomz.
1693 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1694 the src/util/ subdirectory.
1696 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1697 renamed to Transport Post Delivery Action by Jeremy Harris, as
1700 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1701 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1702 redis_servers = needs to be configured which will be used by the redis
1703 lookup. Patch from Warren Baker, of The Packet Hub.
1705 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1707 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1708 hostname or reverse DNS when processing a host list. Used suggestions
1709 from multiple comments on this bug.
1711 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1713 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1714 Missed a few lines, added it to make the runtest require no keyboard
1717 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1718 contains upper case chars. Make router use caseful_local_part.
1720 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1721 support when GnuTLS has been built with p11-kit.
1727 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1729 This, or similar/improved, will also be change PP/11 of 4.82.
1735 PP/01 Handle short writes when writing local log-files.
1736 In practice, only affects FreeBSD (8 onwards).
1737 Bugzilla 1053, with thanks to Dmitry Isaikin.
1739 NM/01 Bugzilla 949 - Documentation tweak
1741 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1744 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1746 PP/02 Implemented gsasl authenticator.
1748 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1750 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1751 `pkg-config foo` for cflags/libs.
1753 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1754 with rest of GSASL and with heimdal_gssapi.
1756 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1757 `pkg-config foo` for cflags/libs for the TLS implementation.
1759 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1760 properties get this fed in as external SSF. A number of robustness
1761 and debugging improvements to the cyrus_sasl authenticator.
1763 PP/08 cyrus_sasl server now expands the server_realm option.
1765 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1766 Patch by Jeremy Harris.
1768 PP/10 Added dbmjz lookup type.
1770 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1772 PP/12 MAIL args handles TAB as well as SP, for better interop with
1773 non-compliant senders.
1774 Analysis and variant patch by Todd Lyons.
1776 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1777 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1778 Patch from Dirk Mueller <dmueller@suse.com>
1780 PP/13 tls_peerdn now print-escaped for spool files.
1781 Observed some $tls_peerdn in wild which contained \n, which resulted
1782 in spool file corruption.
1784 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1785 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1786 or write after TLS renegotiation, which otherwise led to messages
1789 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1790 as a tracking header (ie: a signed header comes before the signature).
1791 Patch from Wolfgang Breyha.
1793 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1794 comma-sep list; embedded commas doubled.
1796 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1798 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1800 Report and patch from Dmitry Banschikov.
1802 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1803 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1804 failures. We appear to now support TLS1.1+ with Exim.
1806 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1807 lets Exim select keys and certificates based upon TLS SNI from client.
1808 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1809 before an outbound SMTP session. New log_selector, +tls_sni.
1811 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1812 NULL dereference. Report and patch from Alun Jones.
1814 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1815 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1816 Not seeing resolver debug output on NetBSD, but suspect this is a
1817 resolver implementation change.
1819 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1820 Left warnings. Added "eximon gdb" invocation mode.
1822 PP/21 Defaulting "accept_8bitmime" to true, not false.
1824 PP/22 Added -bw for inetd wait mode support.
1826 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1827 locate the relevant includes and libraries. Made this the default.
1829 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1830 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1832 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1833 This may cause build issues on older platforms.
1835 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1836 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1837 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1838 Added SNI support via GnuTLS too.
1839 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1841 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1843 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1844 Applied second patch from Janne, implementing suggestion to default
1845 multiple-strings-in-record handling to match SPF spec.
1847 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1849 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1850 read-only, out of scope).
1851 Patch from Wolfgang Breyha, report from Stuart Northfield.
1853 PP/29 Fix three issues highlighted by clang analyser static analysis.
1854 Only crash-plausible issue would require the Cambridge-specific
1855 iplookup router and a misconfiguration.
1856 Report from Marcin Mirosław.
1858 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1860 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1861 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1862 As part of this, removing so much warning spew let me fix some minor
1863 real issues in debug logging.
1865 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1866 assignment on my part. Fixed.
1868 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1869 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1870 Janne Snabb (who went above and beyond: thank you).
1872 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1873 string otherwise requires a connection and a bunch more work and it's
1874 relatively easy to get wrong. Should also expose TLS library linkage
1877 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1878 64-bit ${eval} (JH/03).
1880 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1881 GNU libc to support some of the 64-bit stuff, should not lead to
1882 conflicts. Defined before os.h is pulled in, so if a given platform
1883 needs to override this, it can.
1885 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1886 protection layer was required, which is not implemented.
1887 Bugzilla 1254, patch from Wolfgang Breyha.
1889 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1890 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1891 tls_dhparam take prime identifiers. Also unbreak combination of
1892 OpenSSL+DH_params+TLSSNI.
1894 PP/39 Disable SSLv2 by default in OpenSSL support.
1900 PP/01 Solaris build fix for Oracle's LDAP libraries.
1901 Bugzilla 1109, patch from Stephen Usher.
1903 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1905 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1908 TF/02 Fix a couple more cases where we did not log the error message
1909 when unlink() failed. See also change 4.74-TF/03.
1911 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1912 lock up or crash if it happened to be inside a call to libc when it
1913 got a SIGUSR1 from exiwhat.
1915 The SIGUSR1 handler appends the current process status to the process
1916 log which is later printed by exiwhat. It used to use the general
1917 purpose logging code to do this, but several functions it calls are
1918 not safe for signals.
1920 The new output code in the SIGUSR1 handler is specific to the process
1921 log, and simple enough that it's easy to inspect for signal safety.
1922 Removing some special cases also simplifies the general logging code.
1923 Removing the spurious timestamps from the process log simplifies
1926 TF/04 Improved ratelimit ACL condition.
1928 The /noupdate option has been deprecated in favour of /readonly which
1929 has clearer semantics. The /leaky, /strict, and /readonly update modes
1930 are mutually exclusive. The update mode is no longer included in the
1931 database key; it just determines when the database is updated. (This
1932 means that when you upgrade Exim will forget old rate measurements.)
1934 Exim now checks that the per_* options are used with an update mode that
1935 makes sense for the current ACL. For example, when Exim is processing a
1936 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1937 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1938 must specify per_mail/readonly. If you omit the update mode it defaults to
1939 /leaky where that makes sense (as before) or /readonly where required.
1941 The /noupdate option is now undocumented but still supported for
1942 backwards compatibility. It is equivalent to /readonly except that in
1943 ACLs where /readonly is required you may specify /leaky/noupdate or
1944 /strict/noupdate which are treated the same as /readonly.
1946 A useful new feature is the /count= option. This is a generalization
1947 of the per_byte option, so that you can measure the throughput of other
1948 aggregate values. For example, the per_byte option is now equivalent
1949 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1951 The per_rcpt option has been generalized using the /count= mechanism
1952 (though it's more complicated than the per_byte equivalence). When it is
1953 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1954 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1955 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1956 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1957 Note that using per_rcpt with a non-readonly update mode in more than
1958 one ACL will cause the recipients to be double-counted. (The per_mail
1959 and per_byte options don't have this problem.)
1961 The handling of very low rates has changed slightly. If the computed rate
1962 is less than the event's count (usually one) then this event is the first
1963 after a long gap. In this case the rate is set to the same as this event's
1964 count, so that the first message of a spam run is counted properly.
1966 The major new feature is a mechanism for counting the rate of unique
1967 events. The new per_addr option counts the number of different
1968 recipients that someone has sent messages to in the last time period. It
1969 behaves like per_rcpt if all the recipient addresses are different, but
1970 duplicate recipient addresses do not increase the measured rate. Like
1971 the /count= option this is a general mechanism, so the per_addr option
1972 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1973 example, measure the rate that a client uses different sender addresses
1974 with the options per_mail/unique=$sender_address. There are further
1975 details in the main documentation.
1977 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1979 TF/06 Removed a few PCRE remnants.
1981 TF/07 Automatically extract Exim's version number from tags in the git
1982 repository when doing development or release builds.
1984 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1985 Bugzilla 879. Patch from Paul Fisher.
1987 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1988 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1991 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1993 PP/05 Variable $av_failed, true if the AV scanner deferred.
1994 Bugzilla 1078. Patch from John Horne.
1996 PP/06 Stop make process more reliably on build failure.
1997 Bugzilla 1087. Patch from Heiko Schlittermann.
1999 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2000 Bugzilla 1089. Patch from Heiko Schlittermann.
2002 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2003 Bugzilla 1131. Patch from Holger Weiß.
2005 PP/09 Handle IPv6 addresses with SPF.
2006 Bugzilla 860. Patch from Wolfgang Breyha.
2008 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2010 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2013 PP/11 match_* no longer expand right-hand-side by default.
2014 New compile-time build option, EXPAND_LISTMATCH_RHS.
2015 New expansion conditions, "inlist", "inlisti".
2017 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2019 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2021 PP/14 fix log_write() format string regression from TF/03.
2022 Bugzilla 1152. Patch from Dmitry Isaikin.
2028 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2030 PP/02 Harmonised TLS library version reporting; only show if debugging.
2031 Layout now matches that introduced for other libraries in 4.74 PP/03.
2033 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2035 PP/04 New "dns_use_edns0" global option.
2037 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2040 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2041 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2043 TK/01 Updated PolarSSL code to 0.14.2.
2044 Bugzilla 1097. Patch from Andreas Metzler.
2046 PP/07 Catch divide-by-zero in ${eval:...}.
2047 Fixes bugzilla 1102.
2049 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2052 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2053 format-string attack -- SECURITY: remote arbitrary code execution.
2055 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2056 time unintentionally subject to list matching rules, letting the header
2057 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2058 arbitrary string expansion). This allowed for information disclosure.
2060 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2061 INT_MIN/-1 -- value coerced to INT_MAX.
2067 NM/01 Workaround for PCRE version dependency in version reporting
2070 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2071 This fixes portability to compilers other than gcc, notably
2072 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2074 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2075 makefiles for portability to HP-UX and POSIX correctness.
2077 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2078 Also via indented variable definition in the Makefile.
2079 (Debugging by Oliver Heesakkers).
2081 PP/02 Restore caching of spamd results with expanded spamd_address.
2082 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2084 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2085 Improves build reliability. Fix from: Frank Elsner
2087 NM/02 Fix wide character breakage in the rfc2047 coding
2088 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2090 NM/03 Allow underscore in dnslist lookups
2091 Fixes bug 1026. Patch from Graeme Fowler
2093 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2094 Code patches from Adam Ciarcinski of NetBSD.
2096 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2099 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2100 is logged, to avoid truncation. Patch from John Horne.
2102 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2103 Patch from Jakob Hirsch.
2105 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2106 SQL string expansion failure details.
2107 Patch from Andrey Oktyabrski.
2109 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2110 Patch from Simon Arlott.
2112 PP/09 New lookups functionality failed to compile on old gcc which rejects
2113 extern declarations in function scope.
2114 Patch from Oliver Fleischmann
2116 PP/10 Use sig_atomic_t for flags set from signal handlers.
2117 Check getgroups() return and improve debugging.
2118 Fixed developed for diagnosis in bug 927 (which turned out to be
2121 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2122 Patch from Mark Zealey.
2124 PP/12 Bugzilla 1056: Improved spamd server selection.
2125 Patch from Mark Zealey.
2127 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2128 Based on patch from Heiko Schlittermann.
2130 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2131 Patch from Uwe Doering, sign-off by Michael Haardt.
2133 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2134 variable declaration deep within a block. Bug and patch from
2137 PP/15 lookups-Makefile IRIX compatibility coercion.
2139 PP/16 Make DISABLE_DKIM build knob functional.
2141 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2142 Patch by Simon Arlott
2144 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2145 variable argument macros. Our copy now differs from upstream.
2151 TF/01 Failure to get a lock on a hints database can have serious
2152 consequences so log it to the panic log.
2154 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2155 controlled using the smtp_confirmation log selector.
2157 TF/03 Include the error message when we fail to unlink a spool file.
2159 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2160 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2161 for maintaining out-of-tree patches for some time.
2163 PP/01 Bugzilla 139: Documentation and portability issues.
2164 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2165 Handle per-OS dynamic-module compilation flags.
2167 PP/02 Let /dev/null have normal permissions.
2168 The 4.73 fixes were a little too stringent and complained about the
2169 permissions on /dev/null. Exempt it from some checks.
2170 Reported by Andreas M. Kirchwitz.
2172 PP/03 Report version information for many libraries, including
2173 Exim version information for dynamically loaded libraries. Created
2174 version.h, now support a version extension string for distributors
2175 who patch heavily. Dynamic module ABI change.
2177 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2178 privilege escalation vulnerability whereby the Exim run-time user
2179 can cause root to append content of the attacker's choosing to
2182 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2185 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2186 If dropping privileges for untrusted macros, we disabled normal logging
2187 on the basis that it would fail; for the Exim run-time user, this is not
2188 the case, and it resulted in successful deliveries going unlogged.
2189 Fixed. Reported by Andreas Metzler.
2195 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2196 only prepend for the Resent-* case. Fixes regression introduced in
2197 Exim 4.70 by NM/22 for Bugzilla 607.
2199 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2200 increasing numbers of administrators be bitten by this.
2202 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2204 PP/03 Bugzilla 994: added openssl_options main configuration option.
2206 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2208 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2210 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2212 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2213 without a peer certificate, leading to a segfault because of an
2214 assumption that peers always have certificates. Be a little more
2215 paranoid. Problem reported by Martin Tscholak.
2217 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2218 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2219 NB: ClamAV planning to remove STREAM in "middle of 2010".
2220 CL also introduces -bmalware, various -d+acl logging additions and
2221 more caution in buffer sizes.
2223 PP/09 Implemented reverse_ip expansion operator.
2225 PP/10 Bugzilla 937: provide a "debug" ACL control.
2227 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2229 PP/12 Bugzilla 973: Implement --version.
2231 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2233 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2235 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2237 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2238 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2239 ignore trailing whitespace.
2241 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2243 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2246 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2247 Notification from Dr Andrew Aitchison.
2249 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2250 ExtendedDetectionInfo response format.
2251 Notification from John Horne.
2253 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2256 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2257 XSL and documented dependency on system catalogs, with examples of how
2260 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2263 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2264 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2265 configuration file which is writeable by the Exim user or group.
2267 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2268 of configuration files to cover files specified with the -C option if
2269 they are going to be used with root privileges, not just the default
2272 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2273 option (effectively making it always true).
2275 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2276 files to be used while preserving root privileges.
2278 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2279 that rogue child processes cannot use them.
2281 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2282 run-time user, instead of root.
2284 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2285 Exim run-time user without dropping privileges.
2287 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2288 result string, instead of calling string_vformat() twice with the same
2291 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2292 for other users. Others should always drop root privileges if they use
2293 -C on the command line, even for a whitelisted configure file.
2295 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2297 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2303 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2304 $data_path, and $header_path variables; fixed documentation bugs and
2307 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2308 exipick to access non-standard spools, including the "frozen" queue
2311 NM/01 Bugzilla 965: Support mysql stored procedures.
2312 Patch from Alain Williams
2314 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2316 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2317 Patch from Andreas Metzler
2319 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2320 Patch from Kirill Miazine
2322 NM/05 Bugzilla 671: Added umask to procmail example.
2324 JJ/03 installed exipick 20100323.0, fixing doc bug
2326 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2327 directory. Notification and patch from Dan Rosenberg.
2329 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2331 TK/02 Improve log output when DKIM signing operation fails.
2333 MH/01 Treat the transport option dkim_domain as a colon separated
2334 list, not as a single string, and sign the message with each element,
2335 omitting multiple occurences of the same signer.
2337 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2338 Bugzilla 985, 986. Patch by Simon Arlott
2340 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2341 Patch by Simon Arlott
2343 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2344 MBX locking. Notification from Dan Rosenberg.
2350 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2352 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2354 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2356 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2358 NM/04 Bugzilla 915: Flag broken perl installation during build.
2364 TK/01 Added patch by Johannes Berg that expands the main option
2365 "spamd_address" if it starts with a dollar sign.
2367 TK/02 Write list of recipients to X-Envelope-Sender header when building
2368 the mbox-format spool file for content scanning (suggested by Jakob
2371 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2372 (http://www.dcc-servers.net/) support via dccifd. Activated by
2373 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2375 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2376 by Mark Daniel Reidel <mr@df.eu>.
2378 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2379 When building exim an external PCRE library is now needed -
2380 PCRE is a system library on the majority of modern systems.
2381 See entry on PCRE_LIBS in EDITME file.
2383 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2384 conversation. Added nologin parameter to request.
2385 Patch contributed by Kirill Miazine.
2387 TF/01 Do not log submission mode rewrites if they do not change the address.
2389 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2391 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2392 log files in place. Contributed by Roberto Lima.
2394 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2396 TF/03 Bugzilla 615: When checking the local_parts router precondition
2397 after a local_part_suffix or local_part_prefix option, Exim now
2398 does not use the address's named list lookup cache, since this
2399 contains cached lookups for the whole local part.
2401 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2402 Robert Millan. Documentation is in experimental-spec.txt.
2404 TF/04 Bugzilla 668: Fix parallel build (make -j).
2406 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2408 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2409 Patch provided by Jan Srzednicki.
2411 TF/05 Leading white space used to be stripped from $spam_report which
2412 wrecked the formatting. Now it is preserved.
2414 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2415 that they are available at delivery time.
2417 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2419 TF/08 TLS error reporting now respects the incoming_interface and
2420 incoming_port log selectors.
2422 TF/09 Produce a more useful error message if an SMTP transport's hosts
2423 setting expands to an empty string.
2425 NM/06 Bugzilla 744: EXPN did not work under TLS.
2426 Patch provided by Phil Pennock.
2428 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2429 Patch provided by Richard Godbee.
2431 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2432 acl_smtp_notquit, added index entry.
2434 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2435 Patch provided by Eugene Bujak.
2437 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2438 accept(). Patch provided by Maxim Dounin.
2440 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2441 Patch provided by Phil Pennock.
2443 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2445 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2446 Patch provided by Brad "anomie" Jorsch.
2448 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2449 Patch provided by Dean Brooks.
2451 TK/05 Add native DKIM support (does not depend on external libraries).
2453 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2454 Patch provided by Graeme Fowler.
2456 NM/16 Bugzilla 851: Documentation example syntax fix.
2458 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2460 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2463 NM/19 Bugzilla 745: TLS version reporting.
2464 Patch provided by Phil Pennock.
2466 NM/20 Bugzilla 167: bool: condition support.
2467 Patch provided by Phil Pennock.
2469 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2470 clients. Patch provided by Phil Pennock.
2472 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2473 Patch provided by Brad "anomie" Jorsch.
2475 NM/23 Bugzilla 687: Fix misparses in eximstats.
2476 Patch provided by Heiko Schlittermann.
2478 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2479 Patch provided by Heiko Schlittermann.
2481 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2482 plus update to original patch.
2484 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2486 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2487 Patch provided by David Brownlee.
2489 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2491 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2493 NM/30 Bugzilla 888: TLS documentation bugfixes.
2495 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2497 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2498 Unlike the original bugzilla I have changed all shell scripts in src tree.
2500 NM/33 Bugzilla 898: Transport filter timeout fix.
2501 Patch by Todd Rinaldo.
2503 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2504 Patch by Serge Demonchaux.
2506 NM/35 Bugzilla 39: Base64 decode bug fixes.
2507 Patch by Jakob Hirsch.
2509 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2511 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2513 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2515 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2521 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2522 ALT-N's libdkim that I have put here:
2523 http://duncanthrax.net/exim-experimental/
2525 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2526 were called 'true' and it seems that C99 defines that as a reserved
2527 keyword to be used with 'bool' variable types. That means you could
2528 not include C99-style headers which use bools without triggering
2529 build errors in sieve.c.
2531 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2532 as mailq or other aliases. Changed the --help handling significantly
2533 to do whats expected. exim_usage() emits usage/help information.
2535 SC/01 Added the -bylocaldomain option to eximstats.
2537 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2539 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2541 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2547 PH/01 Another patch from the Sieve maintainer.
2549 PH/02 When an IPv6 address is converted to a string for single-key lookup
2550 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2551 dots are used instead of colons so that keys in lsearch files need not
2552 contain colons. This was done some time before quoting was made available
2553 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2554 (notated using the quote facility) so as to distinguish them from IPv4
2555 keys. This meant that lookups for IP addresses in host lists did not work
2556 for iplsearch lookups.
2558 This has been fixed by arranging for IPv6 addresses to be expressed with
2559 colons if the lookup type is iplsearch. This is not incompatible, because
2560 previously such lookups could never work.
2562 The situation is now rather anomalous, since one *can* have colons in
2563 ordinary lsearch keys. However, making the change in all cases is
2564 incompatible and would probably break a number of configurations.
2566 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2569 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2570 conversion specification without a maximum field width, thereby enabling
2571 a rogue spamd server to cause a buffer overflow. While nobody in their
2572 right mind would setup Exim to query an untrusted spamd server, an
2573 attacker that gains access to a server running spamd could potentially
2574 exploit this vulnerability to run arbitrary code as the Exim user.
2576 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2577 $primary_hostname instead of what libspf2 thinks the hosts name is.
2579 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2580 a directory entry by the name of the lookup key. Previously, if a
2581 symlink pointed to a non-existing file or a file in a directory that
2582 Exim lacked permissions to read, a lookup for a key matching that
2583 symlink would fail. Now it is enough that a matching directory entry
2584 exists, symlink or not. (Bugzilla 503.)
2586 PH/03 The body_linecount and body_zerocount variables are now exported in the
2589 PH/04 Added the $dnslist_matched variable.
2591 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2592 This means they are set thereafter only if the connection becomes
2595 PH/06 Added the client_condition to authenticators so that some can be skipped
2596 by clients under certain conditions.
2598 PH/07 The error message for a badly-placed control=no_multiline_responses left
2599 "_responses" off the end of the name.
2601 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2603 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2604 (without spaces) instead of just copying the configuration text.
2606 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2608 PH/11 Added $max_received_linelength.
2610 PH/12 Added +ignore_defer and +include_defer to host lists.
2612 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2613 way in which PCRE > 7.0 is built.
2615 PH/14 Implemented queue_only_load_latch.
2617 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2618 MAIL command. The effect was to mangle the value on 64-bit systems.
2620 PH/16 Another patch from the Sieve maintainer.
2622 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2624 PH/18 If a system quota error occurred while trying to create the file for
2625 a maildir delivery, the message "Mailbox is full" was not appended to the
2626 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2627 only to a quota excession during the actual writing of the file.
2629 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2630 characters?) which causes problems in log lines. The DN values are now
2631 passed through string_printing() before being added to log lines.
2633 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2634 and InterBase are left for another time.)
2636 PH/21 Added message_body_newlines option.
2638 PH/22 Guard against possible overflow in moan_check_errorcopy().
2640 PH/23 POSIX allows open() to be a macro; guard against that.
2642 PH/24 If the recipient of an error message contained an @ in the local part
2643 (suitably quoted, of course), incorrect values were put in $domain and
2644 $local_part during the evaluation of errors_copy.
2650 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2651 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2654 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2655 issue a MAIL command.
2657 PH/02 In an ACL statement such as
2659 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2661 if a client was not listed at all, or was listed with a value other than
2662 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2663 the condition was not true (as it should be), so access was not denied.
2664 The bug was that the ! inversion was incorrectly passed on to the second
2665 item. This has been fixed.
2667 PH/03 Added additional dnslists conditions == and =& which are different from
2668 = and & when the dns lookup returns more than one IP address.
2670 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2671 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2673 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2674 FSYNC, which compiles an option called disable_fsync that allows for
2675 bypassing fsync(). The documentation is heavily laced with warnings.
2677 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2679 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2680 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2681 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2682 including adding "make clean"; (3) Added -fPIC when compiling the test
2683 dynamically loaded module, to get rid of a warning.
2685 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2686 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2687 = 0s. The bug is otherwise harmless.
2689 PH/07 There was a bug in the dovecot authenticator such that the value of
2690 $auth1 could be overwritten, and so not correctly preserved, after a
2691 successful authentication. This usually meant that the value preserved by
2692 the server_setid option was incorrect.
2694 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2696 PH/09 Installed PCRE release 7.0.
2698 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2699 run for batched SMTP input. It is now run at the start of every message
2700 in the batch. While fixing this I discovered that the process information
2701 (output by running exiwhat) was not always getting set for -bs and -bS
2702 input. This is fixed, and it now also says "batched" for BSMTP.
2704 PH/11 Added control=no_pipelining.
2706 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2707 patch, slightly modified), and move the expansion of helo_data till after
2708 the connection is made in the smtp transport (so it can use these
2711 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2713 PH/14 Added log_selector = +pid.
2715 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2717 PH/16 Add ${if forany and ${if forall.
2719 PH/17 Added dsn_from option to vary the From: line in DSNs.
2721 PH/18 Flush SMTP output before performing a callout, unless control =
2722 no_callout_flush is set.
2724 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2725 was true (the default) a successful delivery failed to delete the retry
2726 item, thus causing premature timeout of the address. The bug is now
2729 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2731 PH/21 Long custom messages for fakedefer and fakereject are now split up
2732 into multiline responses in the same way that messages for "deny" and
2733 other ACL rejections are.
2735 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2736 with slight modification.
2738 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2739 draft, changing the syntax and factoring some duplicate code".
2741 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2742 for deliveries of the second and subsequent messages over the same SMTP
2745 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2746 ${reduce, with only minor "tidies".
2748 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2750 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2751 expansion side effects.
2753 PH/27 When a message times out after an over-quota error from an Exim-imposed
2754 quota, the bounce message says "mailbox is full". This message was not
2755 being given when it was a system quota that was exceeded. It now should
2758 MH/03 Made $recipients available in local_scan(). local_scan() already has
2759 better access to the recipient list through recipients_list[], but
2760 $recipients can be useful in postmaster-provided expansion strings.
2762 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2763 in the case of a MAIL command with additional options following the
2764 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2765 were accidentally chopped off.
2767 PH/29 SMTP synchronization checks are implemented when a command is read -
2768 there is a check that no more input is waiting when there shouldn't be
2769 any. However, for some commands, a delay in an ACL can mean that it is
2770 some time before the response is written. In this time, more input might
2771 arrive, invalidly. So now there are extra checks after an ACL has run for
2772 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2773 pipelining has not been advertised.
2775 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2777 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2778 returned in the SMTP response when smtp_return_error_details was false.
2779 This has been fixed.
2781 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2782 instead of the C I/O that was originally supplied, because problems were
2783 reported on Solaris.
2785 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2786 Exim which did not show up earlier: it was assuming that a call to
2787 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2788 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2789 was a macro that became an assignment, so it seemed to work. This has
2790 changed to a proper function call with a void return, hence the compile
2791 error. Exim's code has been fixed.
2793 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2796 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2798 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2800 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2801 because it turns out to be too big or there is a timeout) there is no
2802 "Completed" line in the log. When some messages of this type were
2803 selected by exigrep, they were listed as "not completed". Others were
2804 picked up by some special patterns. I have improved the selection
2805 criteria to be more general.
2807 PH/38 The host_find_failed option in the manualroute router can now be set
2808 to "ignore", to completely ignore a host whose IP address cannot be
2809 found. If all hosts are ignored, the behaviour is controlled by the new
2810 host_all_ignored option.
2812 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2813 homing or because of multiple MX records with /mx) generated more than
2814 one IP address, and the following item turned out to be the local host,
2815 all the secondary addresses of the first item were incorrectly removed
2816 from the list, along with the local host and any following hosts (which
2817 is what is supposed to happen).
2819 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2820 whoever called Exim into the -H file. In the case of the daemon it was
2821 behaving confusingly. When first started, it used values for whoever
2822 started the daemon, but after a SIGHUP it used the Exim user (because it
2823 calls itself on a restart). I have changed the code so that it now always
2826 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2827 message are rejected with the same error (e.g. no authentication or bad
2828 sender address), and a DATA command is nevertheless sent (as can happen
2829 with PIPELINING or a stupid MUA), the error message that was given to the
2830 RCPT commands is included in the rejection of the DATA command. This is
2831 intended to be helpful for MUAs that show only the final error to their
2834 PH/42 Another patch from the Sieve maintainer.
2836 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2837 Eximstats - Fixed some broken HTML links and added missing column headers
2839 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2840 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2842 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2844 PH/43 Yet another patch from the Sieve maintainer.
2846 PH/44 I found a way to check for a TCP/IP connection going away before sending
2847 the response to the final '.' that terminates a message, but only in the
2848 case where the client has not sent further data following the '.'
2849 (unfortunately, this is allowed). However, in many cases there won't be
2850 any further data because there won't be any more messages to send. A call
2851 to select() can be used: if it shows that the input is "ready", there is
2852 either input waiting, or the socket has been closed. An attempt to read
2853 the next input character can distinguish the two cases. Previously, Exim
2854 would have sent an OK response which the client would never have see.
2855 This could lead to message repetition. This fix should cure that, at
2856 least in a lot of common cases.
2858 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2859 advertised in response to EHLO.
2865 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2866 fixed by 4.65/MH/01 (is this a record?) are fixed:
2868 (i) An empty string was always treated as zero by the numeric comparison
2869 operators. This behaviour has been restored.
2871 (ii) It is documented that the numeric comparison operators always treat
2872 their arguments as decimal numbers. This was broken in that numbers
2873 starting with 0 were being interpreted as octal.
2875 While fixing these problems I realized that there was another issue that
2876 hadn't been noticed. Values of message_size_limit (both the global option
2877 and the transport option) were treated as octal if they started with 0.
2878 The documentation was vague. These values are now always treated as
2879 decimal, and I will make that clear in the documentation.
2885 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2886 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2889 MH/01 Don't check that the operands of numeric comparison operators are
2890 integers when their expansion is in "skipping" mode (fixes bug
2891 introduced by 4.64-PH/07).
2893 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2894 child addresses, Exim now panics and dies. Previously, because the count
2895 is held in a short int, deliveries were likely to be lost. As such a
2896 large number of recipients for a single message is ridiculous
2897 (performance will be very, very poor), I have chosen to impose a limit
2898 rather than extend the field.
2904 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2905 leftover -K file (the existence of which was triggered by #402).
2906 While we were at it, introduced process PID as part of the -K
2907 filename. This should rule out race conditions when creating
2910 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2911 processing considerably. Previous code took too long for large mails,
2912 triggering a timeout which in turn triggers #401.
2914 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2915 in the DK code in transports.c. sendfile() is not really portable,
2916 hence the _LINUX specificness.
2918 TF/01 In the add_headers option to the mail command in an Exim filter,
2919 there was a bug that Exim would claim a syntax error in any
2920 header after the first one which had an odd number of characters
2923 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2924 callout verification, Exim cached a "reject" for the entire domain. This
2925 is correct for most verifications, but it is not correct for a recipient
2926 verification with use_sender or use_postmaster set, because in that case
2927 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2928 case of MAIL FROM:<> rejection from other early rejections (e.g.
2929 rejection of HELO). When verifying a recipient using a non-null MAIL
2930 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2931 Whatever the result of the callout, the value of the domain cache is
2932 left unchanged (for any other kind of callout, getting as far as trying
2933 RCPT means that the domain itself is ok).
2935 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2938 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2939 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2940 session, the daemon ignores ECONNECT errors and logs others; it now
2941 ignores EPIPE as well.
2943 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2944 (quoted-printable decoding).
2946 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2947 later the small subsequent patch to fix an introduced bug.
2949 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2951 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2953 PH/08 An error is now given if message_size_limit is specified negative.
2955 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2956 to be given (somewhat) arbitrary names.
2958 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2961 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2962 miscellaneous code fixes
2964 PH/10 Added the log_reject_target ACL modifier to specify where to log
2967 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2968 hostname. This is wrong, because it relates to the incoming message (and
2969 probably the interface on which it is arriving) and not to the outgoing
2970 callout (which could be using a different interface). This has been
2971 changed to use the value of the helo_data option from the smtp transport
2972 instead - this is what is used when a message is actually being sent. If
2973 there is no remote transport (possible with a router that sets up host
2974 addresses), $smtp_active_hostname is used.
2976 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2977 tweaks were necessary in order to get it to work (see also 21 below):
2978 (a) The code assumed that strncpy() returns a negative number on buffer
2979 overflow, which isn't the case. Replaced with Exim's string_format()
2981 (b) There were several signed/unsigned issues. I just did the minimum
2982 hacking in of casts. There is scope for a larger refactoring.
2983 (c) The code used strcasecmp() which is not a standard C function.
2984 Replaced with Exim's strcmpic() function.
2985 (d) The code set only $1; it now sets $auth1 as well.
2986 (e) A simple test gave the error "authentication client didn't specify
2987 service in request". It would seem that Dovecot has changed its
2988 interface. Fortunately there's a specification; I followed it and
2989 changed what the client sends and it appears to be working now.
2991 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2994 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2995 address A is aliased to B and C, where B exists and C does not. Without
2996 -v the output is "A verified" because verification stops after a
2997 successful redirection if more than one address is generated. However,
2998 with -v the child addresses are also verified. Exim was outputting "A
2999 failed to verify" and then showing the successful verification for C,
3000 with its parentage. It now outputs "B failed to verify", showing B's
3001 parentage before showing the successful verification of C.
3003 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3004 look up a TXT record in a specific list after matching in a combined
3007 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3008 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3009 they consult the DNS. I had assumed they would set it the way they
3010 wanted; and indeed my experiments on Linux seem to show that in some
3011 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3012 To be on the safe side, however, I have now made the interface to
3013 host_find_byname() similar to host_find_bydns(), with an argument
3014 containing the DNS resolver options. The host_find_byname() function now
3015 sets these options at its start, just as host_find_bydns() does. The smtp
3016 transport options dns_qualify_single and dns_search_parents are passed to
3017 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3018 of host_find_byname() use the default settings of RES_DEFNAMES
3019 (qualify_single) but not RES_DNSRCH (search_parents).
3021 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3022 spool_read_header() do less string testing, by means of a preliminary
3023 switch on the second character of optional "-foo" lines. (This is
3024 overdue, caused by the large number of possibilities that now exist.
3025 Originally there were few.) While I was there, I also converted the
3026 str(n)cmp tests so they don't re-test the leading "-" and the first
3027 character, in the hope this might squeeze out yet more improvement.
3029 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3030 flag allowing group syntax was set by the header_syntax check but not
3031 turned off, possible causing trouble later; (2) The flag was not being
3032 set at all for the header_verify test, causing "group"-style headers to
3033 be rejected. I have now set it in this case, and also caused header_
3034 verify to ignore an empty address taken from a group. While doing this, I
3035 came across some other cases where the code for allowing group syntax
3036 while scanning a header line wasn't quite right (mostly, not resetting
3037 the flag correctly in the right place). These bugs could have caused
3038 trouble for malformed header lines. I hope it is now all correct.
3040 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3041 with the "reply" argument non-NULL. The code, however (which originally
3042 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3043 but it didn't always do it. This confused somebody who was copying the
3044 code for some other use. I have removed all the tests.
3046 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3047 feature that was used to support insecure browsers during the U.S. crypto
3048 embargo. It requires special client support, and Exim is probably the
3049 only MTA that supported it -- and would never use it because real RSA is
3050 always available. This code has been removed, because it had the bad
3051 effect of slowing Exim down by computing (never used) parameters for the
3052 RSA_EXPORT functionality.
3054 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3055 authenticator to fail if there's a tab character in the incoming data
3056 (there should never be unless someone is messing about, as it's supposed
3057 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3058 if the connection is using TLS or if the remote IP is the same as the
3059 local IP, and the "valid-client-cert option" if a client certificate has
3062 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3063 authenticators. This can be used for authorization after authentication
3064 succeeds. (In the case of plaintext, it servers for both authentication
3067 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3068 if any retry times were supplied.
3070 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3071 connection, where there is no client IP address to check. In this
3072 situation, the verify now always succeeds.
3074 PH/25 Applied John Jetmore's -Mset patch.
3076 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3078 PH/27 In a string expansion for a processed (not raw) header when multiple
3079 headers of the same name were present, leading whitespace was being
3080 removed from all of them, but trailing whitespace was being removed only
3081 from the last one. Now trailing whitespace is removed from each header
3082 before concatenation. Completely empty headers in a concatenation (as
3083 before) are ignored.
3085 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3086 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3088 PH/29 [Removed. This was a change that I later backed out, and forgot to
3089 correct the ChangeLog entry (that I had efficiently created) before
3090 committing the later change.]
3092 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3093 address errors (4xx response to RCPT) over the same connection as other
3094 messages routed to the same hosts. Such deliveries are always "forced",
3095 so retry times are not inspected. This resulted in far too many retries
3096 for the affected addresses. The effect occurred only when there were more
3097 hosts than the hosts_max_try setting in the smtp transport when it had
3098 the 4xx errors. Those hosts that it had tried were not added to the list
3099 of hosts for which the message was waiting, so if all were tried, there
3100 was no problem. Two fixes have been applied:
3102 (i) If there are any address or message errors in an SMTP delivery, none
3103 of the hosts (tried or untried) are now added to the list of hosts
3104 for which the message is waiting, so the message should not be a
3105 candidate for sending over the same connection that was used for a
3106 successful delivery of some other message. This seems entirely
3107 reasonable: after all the message is NOT "waiting for some host".
3108 This is so "obvious" that I'm not sure why it wasn't done
3109 previously. Hope I haven't missed anything, but it can't do any
3110 harm, as the worst effect is to miss an optimization.
3112 (ii) If, despite (i), such a delivery is accidentally attempted, the
3113 routing retry time is respected, so at least it doesn't keep
3114 hammering the server.
3116 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3117 in ${readsocket because some servers need this prod.
3119 PH/32 Added some extra debug output when updating a wait-xxx database.
3121 PH/33 The hint "could be header name not terminated by colon", which has been
3122 given for certain expansion errors for a long time, was not being given
3123 for the ${if def:h_colon_omitted{... case.
3125 PH/34 The spec says: "With one important exception, whenever a domain list is
3126 being scanned, $domain contains the subject domain." There was at least
3127 one case where this was not true.
3129 PH/35 The error "getsockname() failed: connection reset by peer" was being
3130 written to the panic log as well as the main log, but it isn't really
3131 panic-worthy as it just means the connection died rather early on. I have
3132 removed the panic log writing for the ECONNRESET error when getsockname()
3135 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3136 runs only) independently of the message's sender address. This meant
3137 that, if the 4xx error was in fact related to the sender, a different
3138 message to the same recipient with a different sender could confuse
3139 things. In particular, this can happen when sending to a greylisting
3140 server, but other circumstances could also provoke similar problems.
3141 I have changed the default so that the retry time for these errors is now
3142 based a combination of the sender and recipient addresses. This change
3143 can be overridden by setting address_retry_include_sender=false in the
3146 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3147 remote server are returned as part of bounce messages. This was not
3148 happening for LMTP over a pipe (the lmtp transport), but now it is the
3149 same for both kinds of LMTP.
3151 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3152 in header lines that were in fact CNAMEs. This is no longer the case.
3154 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3155 and queue runs started by the daemon processed all messages. This has
3156 been fixed so that -R and -S can now usefully be given with -q<time>.
3158 PH/40 Import PCRE release 6.7 (fixes some bugs).
3160 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3162 PH/42 Give an error if -q is specified more than once.
3164 PH/43 Renamed the variables $interface_address and $interface_port as
3165 $received_ip_address and $received_port, to make it clear that these
3166 values apply to message reception, and not to the outgoing interface when
3167 a message is delivered. (The old names remain recognized, of course.)
3169 PH/44 There was no timeout on the connect() call when using a Unix domain
3170 socket in the ${readsocket expansion. There now is.
3172 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3173 be meaningful with "accept".
3175 SC/01 Eximstats V1.43
3176 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3178 SC/02 Eximstats V1.44
3179 Use a glob alias rather than an array ref in the generated
3180 parser. This improves both readability and performance.
3182 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3183 Collect SpamAssassin and rejection statistics.
3184 Don't display local sender or destination tables unless
3185 there is data to show.
3186 Added average volumes into the top table text output.
3188 SC/04 Eximstats V1.46
3189 Collect data on the number of addresses (recipients)
3190 as well as the number of messages.
3192 SC/05 Eximstats V1.47
3193 Added 'Message too big' to the list of mail rejection
3194 reasons (thanks to Marco Gaiarin).
3196 SC/06 Eximstats V1.48
3197 Mainlog lines which have GMT offsets and are too short to
3198 have a flag are now skipped.
3200 SC/07 Eximstats V1.49 (Alain Williams)
3201 Added the -emptyok flag.
3203 SC/08 Eximstats V1.50
3204 Fixes for obtaining the IP address from reject messages.
3206 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3207 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3208 whitespace changes from 4.64-PH/27
3210 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3213 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3214 are found, allow negative numbers in numeric criteria)
3216 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3218 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3219 to match changes made in 4.64-PH/43
3221 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3223 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3224 "domain" is actually an IP address, and force a failure. This locks out
3225 those revolvers/nameservers that support "A-for-A" lookups, in
3226 contravention of the specifications.
3228 PH/48 When a host name was looked up from an IP address, and the subsequent
3229 forward lookup of the name timed out, the host name was left in
3230 $sender_host_name, contrary to the specification.
3232 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3233 restricted to single-key lookups, Exim was not diagnosing an error if
3234 * or *@ was used with a query-style lookup.
3236 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3238 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3239 long ago, but noone interested enough thought of it. Let's just say that
3240 the "1.1" means that there are some new functions that weren't there at
3241 some point in the past.
3243 PH/51 Error processing for expansion failure of helo_data from an smtp
3244 transport during callout processing was broken.
3246 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3247 tested/used via the -bh/-bhc/-bs options.
3249 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3250 bug, fixed in subsequent PCRE releases).
3252 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3253 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3255 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3261 SC/01 Use a glob alias rather than an array ref in eximstats generated
3262 parser. This improves both readability and performance.
3264 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3265 Don't display local sender or destination tables in eximstats unless
3266 there is data to show.
3267 Added average volumes into the eximstats top table text output.
3269 SC/03 Collect data on the number of addresses (recipients) as well
3270 as the number of messages in eximstats.
3272 TF/01 Correct an error in the documentation for the redirect router. Exim
3273 does (usually) call initgroups() when daemonizing.
3275 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3276 with consistent privilege compared to when running as a daemon.
3278 TF/03 Note in the spec that $authenticated_id is not set for local
3279 submissions from trusted users.
3281 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3282 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3284 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3285 by adding some example configuration directives to the default
3286 configuration file. A little bit of work is required to uncomment the
3287 directives and define how usernames and passwords are checked, but
3288 there is now a framework to start from.
3290 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3291 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3292 without this. I don't know how relevant this is to other LDAP libraries.
3294 PH/02 Add the verb name to the "unknown ACL verb" error.
3296 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3298 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3300 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3301 directory not expanded when it should be if an expanded home directory
3302 was set for the address (which is overridden by the transport).
3304 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3307 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3308 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3309 because it is too late at that time, and has no effect.
3311 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3312 security issue with \' (bugzilla #107). I could not use the
3313 PQescapeStringConn() function, because it needs a PGconn value as one of
3316 PH/08 When testing addresses using -bt, indicate those final addresses that
3317 are duplicates that would not cause an additional delivery. At least one
3318 person was confused, thinking that -bt output corresponded to deliveries.
3319 (Suppressing duplicates isn't a good idea as you lose the information
3320 about possibly different redirections that led to the duplicates.)
3322 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3323 systems where poll() doesn't work, in particular OS X.
3325 PH/10 Added more information to debugging output for retry time not reached.
3327 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3328 operations in malware.c.
3330 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3333 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3334 syslog_duplication set false, log lines that would normally be written
3335 both the the main log and to the reject log were not written to syslog at
3338 PH/14 In the default configuration, change the use of "message" in ACL warn
3339 statements to "add_header".
3341 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3342 not followed by a command (e.g. "seen endif").
3344 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3345 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3348 PH/17 Added extra conditions to the default value of delay_warning_condition
3352 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3353 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3354 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3357 The Auto-Submitted: and various List- headers are standardised, whereas I
3358 don't think Precedence: ever was.
3360 PH/18 Refactored debugging code in route_finduser() to show more information,
3361 in particular, the error code if getpwnam() issues one.
3363 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3364 This is apparently needed in addition to the PH/07 change above to avoid
3365 any possible encoding problems.
3367 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3368 but not after initializing Perl.
3370 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3371 output them only if debugging. By default they are written stderr,
3372 apparently, which is not desirable.
3374 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3377 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3380 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3382 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3383 authenticated or an ident call has been made. Suppress the default
3384 values for $authenticated_id and $authenticated_sender (but permit -oMai
3385 and -oMas) when testing with -bh.
3387 PH/24 Re-jigged the order of the tests in the default configuration so that the
3388 tests for valid domains and recipients precede the DNS black list and CSA
3389 tests, on the grounds that those ones are more expensive.
3391 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3392 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3393 This bug exists in every version of Exim that I still have, right back to
3396 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3397 However, an attempt to turn on case-sensitivity in a regex key by
3398 including (?-i) didn't work because the subject string was already
3399 lowercased, and the effects were non-intuitive. It turns out that a
3400 one-line patch can be used to allow (?-i) to work as expected.
3406 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3407 other effects) broke the use of negated acl sub-conditions.
3409 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3412 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3413 "Deny" causes Exim to reject the incoming connection with a 554 error.
3414 Unfortunately, if there is a major crisis, such as a disk failure,
3415 tcp-wrappers gives "deny", whereas what one would like would be some
3416 kind of temporary error. A kludge has been added to help with this.
3417 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3418 554 error is used if errno is still zero or contains ENOENT (which occurs
3419 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3422 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3424 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3425 errors. Otherwise a message that provokes a temporary error (when other
3426 messages do not) can cause a whole host to time out.
3428 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3429 addresses were routed directly to files or pipes from a redirect router.
3430 File deliveries just didn't batch; pipe deliveries might have suffered
3433 PH/06 A failure to get a lock for a hints database would erroneously always say
3434 "Failed to get write lock", even when it was really a read lock.
3436 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3437 of 0600. This has been changed to use the value of the lockfile_mode
3438 option (which defaults to 0600).
3440 PH/08 Applied small patch from the Sieve maintainer.
3442 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3443 folder from quota calculations, a direct delivery into this folder messed
3444 up the contents of the maildirsize file. This was because the regex was
3445 used only to exclude .Trash (or whatever) when the size of the mailbox
3446 was calculated. There was no check that a delivery was happening into an
3447 excluded directory. This bug has been fixed by ignoring all quota
3448 processing for deliveries into excluded directories.
3450 PH/10 Added the maildirfolder_create_regex option to appendfile.
3456 PH/01 The code for finding all the local interface addresses on a FreeBSD
3457 system running IPv6 was broken. This may well have applied to all BSD
3458 systems, as well as to others that have similar system calls. The broken
3459 code found IPv4 interfaces correctly, but gave incorrect values for the
3460 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3461 that it would not match correctly against @[] and not recognize the IPv6
3464 PH/02 The ipliteral router was not recognizing addresses of the form user@
3465 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3467 PH/03 Added disable_ipv6.
3469 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3470 decoded form, because it is most often used to construct To: headers
3471 lines in autoreplies, and the decoded form may well be syntactically
3472 invalid. However, $reply_address has leading white space removed, and all
3473 newlines turned into spaces so that the autoreply transport does not
3476 PH/05 If group was specified without a user on a router, and no group or user
3477 was specified on a transport, the group from the router was ignored.
3479 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3480 for visible compile-time settings that can be used to change these
3481 numbers, for those that want even more. Backwards compatibility with old
3482 spool files has been maintained. However, going back to a previous Exim
3483 release will lost any variables that are in spool files.
3485 PH/07 Two small changes when running in the test harness: increase delay when
3486 passing a TCP/IP connection to a new process, in case the original
3487 process has to generate a bounce, and remove special handling of
3488 127.0.0.2 (sic), which is no longer necessary.
3490 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3491 be the same on different OS.
3493 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3496 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3497 whether --show-vars was specified or not
3499 JJ/02 exipick: Added support for new ACL variable spool format introduced
3502 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3503 syntactically invalid From: or Reply-to: line, and a filter used this to
3504 generate an autoreply, and therefore failed to obtain an address for the
3505 autoreply, Exim could try to deliver to a non-existent relative file
3506 name, causing unrelated and misleading errors. What now happens is that
3507 it logs this as a hard delivery error, but does not attempt to create a
3510 PH/11 The exinext utility has a -C option for testing purposes, but although
3511 the given file was scanned by exinext itself; it wasn't being passed on
3512 when Exim was called.
3514 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3515 an end-of-file indication when reading a command response.
3517 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3518 compiled. In many other places in Exim, IPv6 addresses are always
3519 recognized, so I have changed this. It also means that IPv4 domain
3520 literals of the form [IPV4:n.n.n.n] are now always recognized.
3522 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3523 used if the router is not running as root, for example, when verifying at
3524 ACL time, or when using -bh. The debugging output from this situation was
3525 non-existent - all you got was a failure to exec. I have made two
3528 (a) Failures to set uid/gid, the current directory, or a process leader
3529 in a subprocess such as that created by queryprogram now generate
3530 suitable debugging output when -d is set.
3532 (b) The queryprogram router detects when it is not running as root,
3533 outputs suitable debugging information if -d is set, and then runs
3534 the subprocess without attempting to change uid/gid.
3536 PH/15 Minor change to Makefile for building test_host (undocumented testing
3539 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3540 additional section of a DNS packet that returns MX or SRV records.
3541 Instead, it always explicitly searches for A/AAAA records. This avoids
3542 major problems that occur when a DNS server includes only records of one
3543 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3544 fixed another bug: if SRV records were looked up and the corresponding
3545 address records were *not* found in the additional section, the port
3546 values from the SRV records were lost.
3548 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3549 using the correct key (the original address) when searching the retry
3550 rules in order to find which one to use for generating the retry hint.
3552 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3553 adding the default one. Similarly, if it contains a Reply-To: header, the
3554 errors_reply_to option, if set, is not used.
3556 PH/19 When calculating a retry time, Exim used to measure the "time since
3557 failure" by looking at the "first failed" field in the retry record. Now
3558 it does not use this if it is later than than the arrival time of the
3559 message. Instead it uses the arrival time. This makes for better
3560 behaviour in cases where some deliveries succeed, thus re-setting the
3561 "first failed" field. An example is a quota failure for a huge message
3562 when small messages continue to be delivered. Without this change, the
3563 "time since failure" will always be short, possible causing more frequent
3564 delivery attempts for the huge message than are intended.
3565 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3567 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3568 $1, $2, $3) because the numerical variables can be reset during some
3569 expansion items (e.g. "match"), thereby losing the authentication data.
3571 PH/21 Make -bV show the size of off_t variables so that the test suite can
3572 decide whether to run tests for quotas > 2G.
3574 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3575 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3576 filecount value is greater than 2G or if a quota value is greater than 2G
3577 on a system where the size of off_t is not greater than 4, a panic error
3580 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3581 never match. The debug and -bh output now contains an explicit error
3582 message indicating a malformed IPv4 address or mask.
3584 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3585 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3586 PH/23 above applies.
3588 PH/25 Do not write to syslog when running in the test harness. The only
3589 occasion when this arises is a failure to open the main or panic logs
3590 (for which there is an explicit test).
3592 PH/26 Added the /no_tell option to "control=freeze".
3594 PH/27 If a host name lookup failed very early in a connection, for example, if
3595 the IP address matched host_lookup and the reverse lookup yielded a name
3596 that did not have a forward lookup, an error message of the form "no IP
3597 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3598 could be logged. Now it outputs the IP address instead of "NULL".
3600 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3601 allows the sender and the authenticated sender to be set when
3602 submitting a message from within Exim. Since child_open_exim() is
3603 documented for local_scan(), the new function should be too.
3605 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3606 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3607 results in an empty string is now treated as unset.
3609 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3611 PH/31 Added sender_verify_fail logging option.
3613 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3614 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3615 tidied the source and removed it altogether.
3617 PH/33 When a queue run was abandoned because the load average was too high, a
3618 log line was always written; now it is written only if the queue_run log
3619 selector is set. In addition, the log line for abandonment now contains
3620 information about the queue run such as the pid. This is always present
3621 in "start" and "stop" lines but was omitted from the "abandon" line.
3623 PH/34 Omit spaces between a header name and the colon in the error message that
3624 is given when verify = headers_syntax fails (if there are lots of them,
3625 the message gets confusing).
3627 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3628 names, as there are now some PTR records that contain slashes. This check
3629 is only to protect against broken name servers that fall over on strange
3630 characters, so the fact that it applies to all lookups doesn't matter.
3632 PH/36 Now that the new test suite is complete, we can remove some of the
3633 special code in Exim that was needed for the old test suite. For example,
3634 sorting DNS records because real resolvers return them in an arbitrary
3635 order. The new test suite's fake resolver always returns records in the
3638 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3639 bounces) except when queue_only is set, to avoid logging races between
3640 the different processes.
3642 PH/38 Panic-die if .include specifies a non-absolute path.
3644 PH/39 A tweak to the "H" retry rule from its user.
3646 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3647 a label. They prevented compilation on older perls.
3649 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3650 a warning to be raised on newish perls.
3652 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3653 on queue. Changes to match documented behaviour of showing count of
3654 messages matching specified criteria.
3656 PH/40 Changed the default ident timeout from 30s to 5s.
3658 PH/41 Added support for the use of login_cap features, on those BSD systems
3659 that have them, for controlling the resources used by pipe deliveries.
3661 PH/42 The content-scanning code uses fopen() to create files in which to put
3662 message data. Previously it was not paying any attention to the mode of
3663 the files. Exim runs with umask(0) because the rest of the code creates
3664 files with open(), and sets the required mode explicitly. Thus, these
3665 files were ending up world-writeable. This was not a big issue, because,
3666 being within the spool directory, they were not world-accessible. I have
3667 created a function called modefopen, which takes an additional mode
3668 argument. It sets umask(777), creates the file, chmods it to the required
3669 mode, then resets the umask. All the relevant calls to fopen() in the
3670 content scanning code have been changed to use this function.
3672 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3673 to 24 hours. This avoids potential overflow problems when processing G
3674 and H retry rules. I suspect nobody ever tinkers with this value.
3676 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3678 PH/45 When the plaintext authenticator is running as a client, the server's
3679 challenges are checked to ensure they are valid base64 strings. By
3680 default, the authentication attempt is cancelled if an invalid string is
3681 received. Setting client_ignore_invalid_base64 true ignores these errors.
3682 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3683 they are received. Thus, the responses can be made to depend on the
3684 challenges. If an invalid string is ignored, an empty string is placed in
3687 PH/46 Messages that are created by the autoreply transport now contains a
3688 References: header, in accordance with RFCs 2822 and 3834.
3690 PH/47 Added authenticated_sender_force to the smtp transport.
3692 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3694 PH/49 Installed latest patch from the Sieve maintainer.
3696 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3697 was also set, the appendfile transport was unnecessarily scanning a
3698 directory of message files (e.g. for maildir delivery) to find the count
3699 of files (along with the size), even though it did not need this
3700 information. It now does the scan only if it needs to find either the
3701 size of the count of files.
3703 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3705 PH/52 Two bugs concerned with error handling when the smtp transport is
3708 (i) Exim was not creating retry information for temporary errors given
3709 for individual recipients after the DATA command when the smtp transport
3710 was used in LMTP mode. This meant that they could be retried too
3711 frequently, and not timed out correctly.
3713 (ii) Exim was setting the flag that allows error details to be returned
3714 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3715 recipients that were returned after the DATA command.
3717 PH/53 This is related to PH/52, but is more general: for any failing address,
3718 when detailed error information was permitted to be returned to the
3719 sender, but the error was temporary, then after the final timeout, only
3720 "retry timeout exceeded" was returned. Now it returns the full error as
3721 well as "retry timeout exceeded".
3723 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3724 do this, and (what is worse) MTAs that accept it.
3726 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3727 will now be deprecated.
3729 PH/56 New os.c-cygwin from the Cygwin maintainer.
3731 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3732 formats (previously only available in exim formats via -bpr, -bpru,
3733 and -bpra. Now also available in native and exiqgrep formats)
3735 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3736 with very large, slow to parse queues
3738 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3740 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3742 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3743 responses to authentication challenges, though it was showing the
3744 challenges; (ii) I've removed the CR characters from the debug output for
3747 PH/58 Allow for the insertion of a newline as well as a space when a string
3748 is turned into more than one encoded-word during RFC 2047 encoding. The
3749 Sieve code now uses this.
3751 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3752 data_4xx, lost_connection, tls_required.
3754 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3755 message was being sent as an SMTP response.
3757 PH/61 Add -l and -k options to exicyclog.
3759 PH/62 When verifying, if an address was redirected to one new address, so that
3760 verification continued, and the new address failed or deferred after
3761 having set something in $address_data, the value of $address_data was not
3762 passed back to the ACL. This was different to the case when no
3763 redirection occurred. The value is now passed back in both cases.
3765 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3766 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3767 use login_cap.h, so on its own it isn't the distinguishing feature. The
3768 new name refers directly to the setclassresources() function.
3770 PH/65 Added configuration files for NetBSD3.
3772 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3774 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3775 is preferred over IPv4.
3777 PH/68 The bounce_return_message and bounce_return_body options were not being
3778 honoured for bounces generated during the reception of non-SMTP messages.
3779 In particular, this applied to messages rejected by the ACL. This bug has
3780 been fixed. However, if bounce_return_message is true and bounce_return_
3781 body is false, the headers that are returned for a non-SMTP message
3782 include only those that have been read before the error was detected.
3783 (In the case of an ACL rejection, they have all been read.)
3785 PH/69 The HTML version of the specification is now built in a directory called
3786 spec_html instead of spec.html, because the latter looks like a path with
3787 a MIME-type, and this confuses some software.
3789 PH/70 Catch two compiler warnings in sieve.c.
3791 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3792 function verify_get_ident() calls ip_connect() to connect a socket, but
3793 if the "connect()" function timed out, ip_connect() used to close the
3794 socket. However, verify_get_ident() also closes the socket later, and in
3795 between Exim writes to the log, which may get opened at this point. When
3796 the socket was closed in ip_connect(), the log could get the same file
3797 descriptor number as the socket. This naturally causes chaos. The fix is
3798 not to close the socket in ip_connect(); the socket should be closed by
3799 the function that creates it. There was only one place in the code where
3800 this was missing, in the iplookup router, which I don't think anybody now
3801 uses, but I've fixed it anyway.
3803 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3804 well as to direct DNS lookups. Otherwise the handling of names in host
3805 lists is inconsistent and therefore confusing.
3811 PH/01 Two changes to the default runtime configuration:
3813 (1) Move the checks for relay_from_hosts and authenticated clients from
3814 after to before the (commented out) DNS black list checks.
3816 (2) Add control=submission to the relay_from_hosts and authenticated
3817 clients checks, on the grounds that messages accepted by these
3818 statements are most likely to be submissions.
3820 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3822 (1) Generate an error if the third argument for the ${prvs expansion is
3825 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3828 (3) Reset the variables that are obtained from the first argument of
3829 ${prvscheck and used in the second argument before leaving the code,
3830 because their memory is reclaimed, so using them afterwards may do
3833 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3834 one (it's much easier than Tom thought :-).
3836 (5) Because of (4), we can now allow for the use of $prvscheck_result
3837 inside the third argument.
3839 PH/03 For some reason, the default setting of PATH when running a command from
3840 a pipe transport was just "/usr/bin". I have changed it to
3843 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3844 anything to be listed in the output from -bV.
3846 PH/05 When a filter generated an autoreply, the entire To: header line was
3847 quoted in the delivery log line, like this:
3849 => >A.N.Other <ano@some.domain> <original@ddress> ...
3851 This has been changed so that it extracts the operative address. There
3852 may be more than one such address. If so, they are comma-separated, like
3855 => >ano@some.domain,ona@other.domain <original@ddress> ...
3857 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3858 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3859 not being looked up in the rDNS to get a host name, Exim was showing the
3860 IP address twice in Received: lines, even though the IP addresses were
3861 identical. For example:
3863 Received: from [1.2.3.4] (helo=[1.2.3.4])
3865 However, if the real host name was known, it was omitting the HELO data
3866 if it matched the actual IP address. This has been tidied up so that it
3867 doesn't show the same IP address twice.
3869 PH/07 When both +timestamp and +memory debugging was on, the value given by
3870 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3871 called by the debug printing, thereby overwriting the timestamp buffer.
3872 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3874 PH/08 When the original message was included in an autoreply transport, it
3875 always said "this is a copy of the message, including all the headers",
3876 even if body_only or headers_only was set. It now gives an appropriate
3879 PH/09 Applied a patch from the Sieve maintainer which:
3881 o fixes some comments
3882 o adds the (disabled) notify extension core
3883 o adds some debug output for the result of if/elsif tests
3884 o points to the current vacation draft in the documentation
3885 and documents the missing references header update
3889 o fixes a bug in processing the envelope test (when testing
3890 multiple envelope elements, the last element determined the
3893 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3894 Electronic Mail") by including:
3896 Auto-submitted: auto-generated
3898 in the messages that it generates (bounce messages and others, such as
3899 warnings). In the case of bounce messages for non-SMTP messages, there was
3900 also a typo: it was using "Auto_submitted" (underscore instead of
3901 hyphen). Since every message generated by Exim is necessarily in response
3902 to another message, thes have all been changed to:
3904 Auto-Submitted: auto-replied
3906 in accordance with these statements in the RFC:
3908 The auto-replied keyword:
3910 - SHOULD be used on messages sent in direct response to another
3911 message by an automatic process,
3913 - MUST NOT be used on manually-generated messages,
3915 - MAY be used on Delivery Status Notifications (DSNs) and Message
3916 Disposition Notifications (MDNs),
3918 - MUST NOT be used on messages generated by automatic or periodic
3919 processes, except for messages which are automatic responses to
3922 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3923 to the default Received: header definition.
3925 PH/12 Added log selector acl_warn_skipped (default on).
3927 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3928 variables because (a) they are in the wrong storage pool and (b) even if
3929 they were copied, it wouldn't work properly because of the caching.
3931 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3932 checking when decoding. Apparently there are clients that generate
3933 overlong encoded strings. Why am I not surprised?
3935 PH/15 If the first argument of "${if match_address" was not empty, but did not
3936 contain an "@" character, Exim crashed. Now it writes a panic log message
3937 and treats the condition as false.
3939 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3941 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3942 extension "envelope-auth". The code is finished and in agreement with
3943 other implementations, but there is no documentation so far and in fact,
3944 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3945 not changing the active code.
3947 Print executed "if" and "elsif" statements when debugging is used. This
3948 helps a great deal to understand what a filter does.
3950 Document more things not specified clearly in RFC3028. I had all this
3951 sorted out, when out of a sudden new issues came to my mind. Oops."
3953 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3956 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3957 canonical form (as documented). However, after a host name lookup from
3958 the IP address, check_host() was doing a simple string comparison with
3959 addresses acquired from the DNS when checking that the found name did
3960 have the original IP as one of its addresses. Since any found IPv6
3961 addresses are likely to be in abbreviated form, the comparison could
3962 fail. Luckily, there already exists a function for doing the comparison
3963 by converting both addresses to binary, so now that is used instead of
3964 the text comparison.
3966 PH/20 There was another similar case to PH/19, when a complete host name was
3967 given in a host list; looking up its IP address could give an abbreviated
3968 form, whereas the current host's name might or might not be abbreviated.
3969 The same fix has been applied.
3975 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3976 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3979 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3980 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3982 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3984 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3985 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3986 Also, for "Auto-Submitted", treat anything other than "no" as
3987 non-personal, in accordance with RFC 3834. (Previously it treated
3988 anything starting "auto-" as non-personal.)
3990 TF/01 The control=submission/name=... option had a problem with syntax
3991 errors if the name included a slash character. The /name= option
3992 now slurps the rest of the string, so it can include any characters
3993 but it must come last in the list of options (after /sender_retain
3996 PH/05 Some modifications to the interface to the fake nameserver for the new
4004 TK/01 Added the "success_on_redirect" address verification option. See
4005 NewStuff for rationale and an example.
4007 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4009 PH/02 Patch to exigrep to allow it to work on syslog lines.
4011 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4012 fread() to skip over the body file's header line, because in Cygwin the
4013 header line is locked and is inaccessible.
4015 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4016 co-exist for some time) to make it clear that it is the Exim ID that is
4017 referenced, not the Message-ID: header line.
4019 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4020 string_format() function, because snprintf() does not exist on all
4023 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4024 ${stat: expansion item.
4026 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4027 protocol synchronization error", to keep the pedants happy.
4029 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4030 well as for IRIX systems, when gcc is being used. See the host.c source
4033 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4035 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4038 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4039 transport and to the smtp transport in LMTP mode.
4041 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4043 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4044 run a filter in a subprocess. This could lead to confusion in subsequent
4045 lookups in the parent process. There should also be a search_tidyup() at
4046 the end of the subprocess.
4048 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4049 only if the host matched helo_try_verify_hosts, which caused the
4050 verification to occur when the EHLO/HELO command was issued. The ACL just
4051 tested the remembered result. Now, if a previous verification attempt has
4052 not happened, "verify = helo" does it there and then.
4054 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4056 TK/03 Fix log output including CR from clamd.
4058 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4059 exist provoked a memory error which could cause a segfault.
4061 PH/15 Installed PCRE 6.2
4063 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4065 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4066 of the problem. Specifically, suggested +O2 rather than +O1 for the
4069 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4071 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4072 record, the port was not correctly logged when the outgoing_port log
4073 selector was set (it logged the transort's default port).
4075 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4076 fallback_hosts, and "hosts" in the smtp transport.
4078 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4079 host errors such as "Connection refused".
4081 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4082 authentication with radiusclient 0.4.9:
4084 - Error returned from rc_read_config was caught wrongly
4085 - Username/password not passed on to radius server due to wrong length.
4087 The presumption is that some radiusclient API changes for 4.51/PH/17
4088 were not taken care of correctly. The code is still untested by me (my
4089 Linux distribution still has 0.3.2 of radiusclient), but it was
4090 contributed by a Radius user.
4092 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4093 expanding the "port" option of the smtp transport.
4095 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4096 while reading a MIME header. Thanks to Tom Hughes for a patch.
4098 PH/24 Include config.h inside local_scan.h so that configuration settings are
4101 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4102 that in an ACL for RCPT (for example), you can examine exactly what was
4105 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4106 commands, but it was not correctly comparing the address with the actual
4107 client host address. Thus, it would show the EHLO address in Received:
4108 header lines when this was not necessary.
4110 PH/27 Added the % operator to ${eval:}.
4112 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4113 it should be ignoring failures (because with -C, for example, it has lost
4114 privilege). It wasn't ignoring creation failures other than "already
4117 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4120 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4121 because an input line was too long, either on its own, or by virtue of
4122 too many continuations, the temporary file was not being removed, and the
4123 return code was incorrect.
4125 PH/31 Missing "BOOL" in function definition in filtertest.c.
4127 PH/32 Applied Sieve patches from the maintainer.
4129 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4131 PH/33 Added "verify = not_blind".
4133 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4134 Local/Makefile (with some defaults set). These are used in built scripts
4135 such as exicyclog, but they have never been used in the exim_install
4136 script (though there are many overriding facilities there). I have
4137 arranged that the exim_install script now takes note of these two
4140 PH/35 Installed configuration files for Dragonfly.
4142 PH/36 When a locally submitted message by a trusted user did not contain a
4143 From: header, and the sender address was obtained from -f or from an SMTP
4144 MAIL command, and the trusted user did not use -F to supply a sender
4145 name, $originator_name was incorrectly used when constructing a From:
4146 header. Furthermore, $originator_name was used for submission mode
4147 messages from external hosts without From: headers in a similar way,
4148 which is clearly wrong.
4150 PH/37 Added control=suppress_local_fixups.
4152 PH/38 When log_selector = +received_sender was set, and the addition of the
4153 sender made the log line's construction buffer exactly full, or one byte
4154 less than full, an overflow happened when the terminating "\n" was
4157 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4158 when the result of a list match is failure because a DNS lookup failed.
4160 PH/40 RM_COMMAND is now used in the building process.
4162 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4163 the "build-* directories that it finds.
4165 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4166 address in a domain literal was a prefix of an interface address.
4168 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4169 when verifying a sender address, unless rewrite_headers is false.
4171 PH/44 Wrote a long comment about why errors_to addresses are verified as
4172 recipients, not senders.
4174 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4175 the ratelimit ACL was added.
4177 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4179 PH/46 Added extra information about PostgreSQL errors to the error string.
4181 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4182 avoiding the need to install special zones in a real server. This is
4183 backwards compatible; if it can't find the fake resolver, it drops back.
4184 Thus, both old and new test suites can be run.
4186 TF/02 Added util/ratelimit.pl
4188 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4189 clock is set back in time.
4191 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4192 Candler <B.Candler@pobox.com>.
4194 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4195 OK for addresses that are the result of redirections.
4197 PH/48 A number of further additions for the benefit of the new test suite,
4198 including a fake gethostbyname() that interfaces to the fake DNS resolver
4201 TF/06 The fix for widen_domains has also been applied to qualify_single and
4202 search_parents which are the other dnslookup options that can cause
4205 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4208 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4210 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4211 service" if it encountered an MX record with an empty target hostname.
4212 The message is now "an MX or SRV record indicated no SMTP service".
4214 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4215 if the DNS of the sending site is misconfigured. This is quite a
4216 common situation. This change restores the behaviour of treating a
4217 helo verification defer as a failure.
4219 PH/51 If self=fail was set on a router, the bounce message did not include the
4220 actual error message.
4226 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4228 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4229 command itself ended in error, the underlying message about the transport
4230 filter timeout was being overwritten with the pipe command error. Now the
4231 underlying error message should be appended to the second error message.
4233 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4235 PH/02 Reduce the amount of output that "make" produces by default. Full output
4236 can still be requested.
4238 PH/03 The warning log line about a condition test deferring for a "warn" verb
4239 was being output only once per connection, rather than after each
4240 occurrence (because it was using the same function as for successful
4241 "warn" verbs). This seems wrong, so I have changed it.
4243 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4244 it should not have, which might have caused a crash in the right
4245 circumstances, but probably never did.
4247 PH/04 Installed a modified version of Tony Finch's patch to make submission
4248 mode fix the return path as well as the Sender: header line, and to
4249 add a /name= option so that you can make the user's friendly name appear
4252 TF/03 Added the control = fakedefer ACL modifier.
4254 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4255 Mark Lowes for thorough testing.
4257 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4259 TK/03 Merged latest SRS patch from Miles Wilton.
4261 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4262 with the definition in sysexits.h (which is #included earlier).
4263 Fortunately, Exim does not actually use EX_OK. The code used to try to
4264 preserve the sysexits.h value, by assuming that macro definitions were
4265 scanned for macro replacements. I have been disabused of this notion,
4266 so now the code just undefines EX_OK before #including unistd.h.
4268 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4269 in the smtp transport. When a block could not be written in a single
4270 write() function, the timeout was being re-applied to each part-write.
4271 This seems wrong - if the receiver was accepting one byte at a time it
4272 would take for ever. The timeout is now adjusted when this happens. It
4273 doesn't have to be particularly precise.
4275 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4276 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4278 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4279 without a domain if the check to <postmaster@domain> fails.
4281 SC/01 Eximstats: added -xls and the ability to specify output files
4282 (patch written by Frank Heydlauf).
4284 SC/02 Eximstats: use FileHandles for outputting results.
4286 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4288 SC/04 Eximstats: fixed display of large numbers with -nvr option
4290 SC/05 Eximstats: fixed merging of reports with empty tables.
4292 SC/06 Eximstats: added the -include_original_destination flag
4294 SC/07 Eximstats: removed tabs and trailing whitespace.
4296 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4298 TK/06 MBOX spool code: Add real "From " MBOX separator line
4299 so the .eml file is really in mbox format (even though
4300 most programs do not really care). Patch from Alex Miller.
4302 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4303 The latter is generated from $received_to and is only set if the
4304 message has one envelope recipient. SA can use these headers,
4305 obviously out-of-the-box. Patch from Alex Miller.
4307 PH/08 The ${def test on a variable was returning false if the variable's
4308 value was "0", contrary to what the specification has always said!
4309 The result should be true unless the variable is empty.
4311 PH/09 The syntax error of a character other than { following "${if
4312 def:variable_name" (after optional whitespace) was not being diagnosed.
4313 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4314 accidental colon was present, for example, could give incorrect results.
4316 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4317 result is used (not including appendfile, where other changes are about
4320 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4321 This involved changing a lot of size variables from int to off_t. It
4322 should work with maildirs and everything.
4324 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4325 spamd dying while we are connected to it.
4327 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4330 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4331 function for BDB 4.3.
4333 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4335 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4336 the sources for PCRE have been re-arranged and more files are now
4339 PH/15 The code I had for printing potentially long long variables in PH/11
4340 above was not the best (it lost precision). The length of off_t variables
4341 is now inspected at build time, and an appropriate printing format (%ld
4342 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4343 to be "long long int" or "long int". This is needed for the internal
4344 formatting function string_vformat().
4346 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4347 the configuration file to be ":syslog", then the script "guesses" where
4348 the logs files are, rather than using the compiled in default. In our
4349 case the guess is not the same as the compiled default, so the script
4350 suddenly stopped working when I started to use syslog. The patch checks
4351 to see if log_file_path is "". If so, it attempts to read it from exim
4352 with no configuration file to get the compiled in version, before it
4353 falls back to the previous guessing code."
4355 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4356 implementing BATV in an Exim configuration. See NewStuff for the gory
4359 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4360 Makefile that are specific to HP-UX.
4362 PH/18 If the "use_postmaster" option was set for a recipient callout together
4363 with the "random" option, the postmaster address was used as the MAIL
4364 FROM address for the random test, but not for the subsequent recipient
4365 test. It is now used for both.
4367 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4368 patch removes a few documentation additions to RFC 3028, because the
4369 latest draft now contains them. It adds the new en;ascii-case comparator
4370 and a new error check for 8bit text in MIME parts. Comparator and
4371 require names are now matched exactly. I enabled the subaddress
4372 extension, but it is not well tested yet (read: it works for me)."
4374 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4375 rework some of the code of TK/09 above to avoid the hardwired use of
4376 "%lld" and "long long". Replaced the call to snprintf() with a call to
4379 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4380 records point to non-existent hosts", "retry timeout exceeded", and
4381 "retry time not reached for any host after a long failure period".
4383 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4384 experimental DomainKeys support:
4386 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4387 (2) On an error such as an illegally used "control", the wrong name for
4388 the control was given.
4390 These problems did NOT occur unless DomainKeys support was compiled.
4392 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4394 PH/24 Added ${if match_ip condition.
4396 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4397 databases so that it will be absolutely obvious if a crash occurs in the
4398 DB library. This is a regular occurrence (often caused by mis-matched
4401 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4402 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4403 picked up on a user's system that detects such things. There doesn't seem
4404 to be a gcc warning option for this - only an attribute that has to be
4405 put on the function's prototype. It seems that in Fedora Core 4 they have
4406 set this on a number of new functions. No doubt there will be more in due
4409 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4410 specify a transport. However, if an address that was verified by such a
4411 router was the subject of a callout, Exim crashed because it tried to
4412 read the rcpt_include_affixes from the non-existent transport. Now it
4413 just assumes that the setting of that option is false. This bug was
4414 introduced by 4.51/PH/31.
4416 PH/28 Changed -d+all to exclude +memory, because that information is very
4417 rarely of interest, but it makes the output a lot bigger. People tend to
4418 do -d+all out of habit.
4420 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4421 code in os-type was giving problems when libc.so lives in lib64, like on
4424 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4425 aren't the modern standard, and it seems that some systems' include files
4426 don't always have them. Exim was already checking for some of the newer
4427 ones like T_AAAA, and defining it itself. I've added checks for all the
4428 record types that Exim uses.
4430 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4431 not automatically generating a new one, as it is supposed to. This
4432 prevented TLS from working. If the file did exist, but contained invalid
4433 data, a new version was generated, as expected. It was only the case of a
4434 non-existent file that was broken.
4436 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4437 with a change in libdomainkeys > 0.64.
4439 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4440 from DNS. If the selector record carries the flag, it now has
4441 precedence over the domain-wide flag.
4443 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4445 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4446 the use of a port name that isn't defined in /etc/services) occurred, the
4447 message was deferred as in a normal delivery, and thus remained on the
4448 spool, instead of being failed because of the mua_wrapper setting. This
4449 is now fixed, and I tidied up some of the mua_wrapper messages at the
4452 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4453 the messages in a hash of arrays rather than using individual hashes.
4454 This is a bit cleaner and results in dramatic memory savings, albeit
4455 at a slight CPU cost.
4457 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4458 as requested by Marc Sherman.
4460 SC/10 Eximstats: added histograms for user specified patterns as requested
4463 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4465 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4466 fopen() in the content-scanning modules that did not already have it.
4472 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4473 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4475 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4477 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4479 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4480 to test Sieve filters that use "vacation".
4482 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4483 that changes the way the GnuTLS parameters are stored in the cache file.
4484 The new format can be generated externally. For backward compatibility,
4485 if the data in the cache doesn't make sense, Exim assumes it has read an
4486 old-format file, and it generates new data and writes a new file. This
4487 means that you can't go back to an older release without removing the
4490 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4491 work if there are any delivery delays because "one_time" forces the
4492 parent to be marked "delivered", so its unseen clone is never tried
4493 again. For this reason, Exim now forbids the simultaneous setting of
4496 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4497 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4498 entry if you want to know the details. The fix, however, neglected to
4499 consider the case when local delivery batching is involved. The test for
4500 "previously delivered" was not happening when checking to see if an
4501 address could be batched with a previous (undelivered) one; under
4502 certain circumstances this could lead to multiple deliveries to the same
4505 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4506 in its include files, and this causes problems building Exim.
4508 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4509 header_syntax) but Exim was just ignoring anything given after a slash.
4510 In particular, this caused confusion with an attempt to use "verify =
4511 reverse_host_lookup/defer_ok". An error is now given when options are
4512 supplied for verify items that do not have them. (Maybe reverse_host_
4513 lookup should have a defer_ok option, but that's a different point.)
4515 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4516 defined by RFC 821) to 2048, because there were problems with some AUTH
4517 commands, and RFC 1869 says the size should be increased for extended
4518 SMTP commands that take arguments.
4520 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4523 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4524 "unknown" error; now it says that the functionality isn't in the binary.
4526 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4527 an address' error message when a string expansion fails (syntax or
4528 whatever). Otherwise the password may appear in the log. Following change
4529 PH/42 below, there is no longer a chance of it appearing in a bounce
4532 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4534 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4535 of its addresses was ever tried. (Bugzilla bug #2.)
4537 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4538 the result incorrectly in the debug output. (It correctly added a newline
4539 to what was transported.)
4541 TF/01 Added $received_time.
4543 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4544 commented out examples of how to interface to a virus scanner and to
4545 SpamAssassin. Also added commented examples of av_scanner and
4546 spamd_address settings.
4548 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4549 and controls are allowed in which ACLs. There were a couple of minor
4550 errors. Some of the entries in the conditions table (which is a table of
4551 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4552 negation of where the condition IS allowed.
4554 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4556 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4557 header file does not have a version number, so I've had to invent a new
4558 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4559 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4560 radiusclient), but it was contributed by a Radius user.
4562 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4563 files or directories, for OpenSSL.
4565 PH/19 When an Exim process that is running as root has to create an Exim log
4566 file, it does so in a subprocess that runs as exim:exim so as to get the
4567 ownership right at creation (otherwise, other Exim processes might see
4568 the file with the wrong ownership). There was no test for failure of this
4569 fork() call, which would lead to the process getting stuck as it waited
4570 for a non-existent subprocess. Forks do occasionally fail when resources
4571 run out. I reviewed all the other calls to fork(); they all seem to check
4574 PH/20 When checking for unexpected SMTP input at connect time (before writing
4575 the banner), Exim was not dealing correctly with a non-positive return
4576 from the read() function. If the client had disconnected by this time,
4577 the result was a log entry for a synchronization error with an empty
4578 string after "input=" when read() returned zero. If read() returned -1
4579 (an event I could not check), uninitialized data bytes were printed.
4580 There were reports of junk text (parts of files, etc) appearing after
4583 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4585 PH/22 Added support for macro redefinition, and (re)definition in between
4586 driver and ACL definitions.
4588 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4589 forgetting to use the resulting value; it was using the unexpanded value.
4591 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4592 hadn't been configured. The fix is from Juergen Kreileder, who
4593 understands it better than I do:
4595 "Here's what I see happening with three configured cyrus_sasl
4596 authenticators configured (plain, login, cram-md5):
4598 On startup auth_cyrus_sasl_init() gets called for each of these.
4599 This means three calls to sasl_listmech() without a specified mech_list.
4600 => SASL tests which mechs of all available mechs actually work
4601 => three warnings about OTP not working
4602 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4604 With the patch, sasl_listmech() also gets called three times. But now
4605 SASL's mech_list option is set to the server_mech specified in the the
4606 authenticator. Or in other words, the answer from sasl_listmech()
4607 gets limited to just the mech you're testing for (which is different
4609 => the return list contains just 'plain' or 'login', 'cram-md5' or
4610 nothing depending on the value of ob->server_mech.
4612 I've just tested the patch: Authentication still works fine,
4613 unavailable mechs specified in the exim configuration are still
4614 caught, and the auth.log warnings about OTP are gone."
4616 PH/25 When debugging is enabled, the contents of the command line are added
4617 to the debugging output, even when log_selector=+arguments is not
4620 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4621 answer is "GNU", and only if the return is "GNU/something" is the answer
4624 PH/27 $acl_verify_message is now set immediately after the failure of a
4625 verification in an ACL, and so is available in subsequent modifiers. In
4626 particular, the message can be preserved by coding like this:
4628 warn !verify = sender
4629 set acl_m0 = $acl_verify_message
4631 Previously, $acl_verify_message was set only while expanding "message"
4632 and "log_message" when a very denied access.
4634 PH/28 Modified OS/os.c-Linux with
4636 -#ifndef OS_LOAD_AVERAGE
4637 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4639 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4640 nomenclature these days.)
4642 PH/29 Installed patch from the Sieve maintainer that adds the options
4643 sieve_useraddress and sieve_subaddress to the redirect router.
4645 PH/30 In these circumstances:
4646 . Two addresses routed to the same list of hosts;
4647 . First host does not offer TLS;
4648 . First host accepts first address;
4649 . First host gives temporary error to second address;
4650 . Second host offers TLS and a TLS session is established;
4651 . Second host accepts second address.
4652 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4653 and peerdn, if requested) that were in fact used only for the second
4656 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4657 attention to any local part prefix or suffix that was matched by the
4658 router that accepted the address. It now behaves in the same way as it
4659 does for delivery: the affixes are removed from the local part unless
4660 rcpt_include_affixes is set on the transport.
4662 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4663 timeout during the DATA phase of an incoming message.
4665 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4666 applied a patch sanctioned by the Sieve maintainer.
4668 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4669 the uid or gid is negative. A case of a negative gid caused this to be
4670 noticed. The fix allows for either to be negative.
4672 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4673 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4676 PH/36 The change PH/12 above was broken. Fixed it.
4678 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4679 the grounds that routing the same address twice would always produce the
4680 same answer. This might have been true once, but it is certainly no
4681 longer true now. Routing a child address may depend on the previous
4682 routing that produced that child. Some complicated redirection strategies
4683 went wrong when messages had multiple recipients, and made Exim's
4684 behaviour dependent on the order in which the addresses were given.
4686 I have moved the duplicate checking until after the routing is complete.
4687 Exim scans the addresses that are assigned to local and remote
4688 transports, and removes any duplicates. This means that more work will be
4689 done, as duplicates will always all be routed, but duplicates are
4690 presumably rare, so I don't expect this is of any significance.
4692 For deliveries to pipes, files, and autoreplies, the duplicate checking
4693 still happens during the routing process, since they are not going to be
4696 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4697 It corrects a timeout issue with spamd. This is Ian's comment: "The
4698 background is that sometimes spamd either never reads data from a
4699 connection it has accepted, or it never writes response data. The exiscan
4700 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4701 blindly assumes that writes won't block so it may never time out."
4703 PH/39 Allow G after quota size as well as K and M.
4705 PH/40 The value set for $authenticated_id in an authenticator may not contain
4706 binary zeroes or newlines because the value is written to log lines and
4707 to spool files. There was no check on this. Now the value is run through
4708 the string_printing() function so that such characters are converted to
4709 printable escape sequences.
4711 PH/41 $message_linecount is a new variable that contains the total number of
4712 lines in the message. Compare $body_linecount, which is the count for the
4715 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4716 bounce and delay warning messages, except in certain special cases, which
4719 (a) An SMTP error message from a remote host;
4720 (b) A message specified in a :fail: redirection;
4721 (c) A message specified in a "fail" command in a system filter;
4722 (d) A message specified in a FAIL return from the queryprogram router;
4723 (e) A message specified by the cannot_route_message router option.
4725 In these cases only, Exim does include the error details in bounce and
4726 warning messages. There are also a few cases where bland messages such
4727 as "unrouteable address" or "local delivery error" are given.
4729 PH/43 $value is now also set for the "else" part of a ${run expansion.
4731 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4732 being worked on, but at least Exim now implements the latest version to
4735 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4736 process to complete was treated as a delivery failure, a timeout while
4737 writing the message to the pipe was logged, but erroneously treated as a
4738 successful delivery. Such timeouts include transport filter timeouts. For
4739 consistency with the overall process timeout, these timeouts are now
4740 treated as errors, giving rise to delivery failures by default. However,
4741 there is now a new Boolean option for the pipe transport called
4742 timeout_defer, which, if set TRUE, converts the failures into defers for
4743 both kinds of timeout. A transport filter timeout is now identified in
4746 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4747 systems where "make" and "gmake" are different, calling "gmake" at top
4748 level broke things. I've arranged for the value of $(MAKE) to be passed
4749 from the Makefile to this script so that it can call the same version of
4753 A note about Exim versions 4.44 and 4.50
4754 ----------------------------------------
4756 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4757 changes of various kinds. As a consequence, a big documentation update was
4758 needed. This delayed the release for rather longer than seemed good, especially
4759 in the light of a couple of (minor) security issues. Therefore, the changes
4760 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4761 release. So 4.44 and 4.50 are in effect two different branches that both start
4764 I have left the 4.50 change log unchanged; it contains all the changes since
4765 4.43. The change log for 4.44 is below; many of its items are identical to
4766 those for 4.50. This seems to be the most sensible way to preserve the
4767 historical information.
4773 1. Minor wording change to the doc/README.SIEVE file.
4775 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4776 computation of the current number of files was incorrect.
4778 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4779 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4782 4. Give more explanation in the error message when the command for a transport
4783 filter fails to execute.
4785 5. There are several places where Exim runs a non-Exim command in a
4786 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4787 was being done only for the command run by the queryprogram router. It is
4788 now done for all such subprocesses. The other cases are: ${run, transport
4789 filters, and the commands run by the lmtp and pipe transports.
4791 6. Added CONFIGURE_GROUP build-time option.
4793 7. Some older OS have a limit of 256 on the maximum number of file
4794 descriptors. Exim was using setrlimit() to set 1000 as a large value
4795 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4796 systems. I've change it so that if it can't get 1000, it tries for 256.
4798 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4799 was an oversight, and furthermore, ever since the addition of extra
4800 controls (e.g. 4.43/32), the checks on when to allow different forms of
4801 "control" were broken. There should now be diagnostics for all cases when a
4802 control that does not make sense is encountered.
4804 9. Added the /retain_sender option to "control=submission".
4806 10. $recipients is now available in the predata ACL (oversight).
4808 11. Tidy the search cache before the fork to do a delivery from a message
4809 received from the command line. Otherwise the child will trigger a lookup
4810 failure and thereby defer the delivery if it tries to use (for example) a
4811 cached ldap connection that the parent has called unbind on.
4813 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4814 of $address_data from the recipient verification was clobbered by the
4815 sender verification.
4817 13. The value of address_data from a sender verification is now available in
4818 $sender_address_data in subsequent conditions in the ACL statement.
4820 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4822 15. Added a new option "connect=<time>" to callout options, to set a different
4825 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4826 was its contents. (It was OK if the option was not defined at all.)
4828 17. A "Completed" log line is now written for messages that are removed from
4829 the spool by the -Mrm option.
4831 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4832 information about exactly what failed.
4834 19. Added -dd to debug only the daemon process.
4836 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4837 handles timeouts, both on the server side and network timeouts. Renamed the
4838 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4840 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4841 It is now set to "smtps".
4843 22. $host_address is now set to the target address during the checking of
4844 ignore_target_hosts.
4846 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4847 being passed; this would have caused $sender_host_name to have been used if
4848 matching the list had actually called for a host name (not very likely,
4849 since this list is usually IP addresses). A host name is now passed as
4852 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4853 code that specifies a non-restarting handler (typically sigaction() in
4854 modern systems) in an attempt to fix a rare and obscure crash bug.
4856 25. Narrowed the window for a race in the daemon that could cause it to ignore
4857 SIGCHLD signals. This is not a major problem, because they are used only to
4858 wake it up if nothing else does.
4860 26. A malformed maildirsize file could cause Exim to calculate negative values
4861 for the mailbox size or file count. Odd effects could occur as a result.
4862 The maildirsize information is now recalculated if the size or filecount
4865 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4866 support for a long time. Removed HAVE_SYS_VFS_H.
4868 28. Installed the latest version of exipick from John Jetmore.
4870 29. In an address list, if the pattern was not a regular expression, an empty
4871 subject address (from a bounce message) matched only if the pattern was an
4872 empty string. Non-empty patterns were not even tested. This was the wrong
4873 because it is perfectly reasonable to use an empty address as part of a
4874 database query. An empty address is now tested by patterns that are
4875 lookups. However, all the other forms of pattern expect the subject to
4876 contain a local part and a domain, and therefore, for them, an empty
4877 address still always fails if the pattern is not itself empty.
4879 30. Exim went into a mad DNS loop when attempting to do a callout where the
4880 host was specified on an smtp transport, and looking it up yielded more
4881 than one IP address.
4883 31. Re-factored the code for checking spool and log partition space into a
4884 function that finds that data and another that does the check. The former
4885 is then used to implement four new variables: $spool_space, $log_space,
4886 $spool_inodes, and $log_inodes.
4888 32. The RFC2047 encoding function was originally intended for short strings
4889 such as real names; it was not keeping to the 75-character limit for
4890 encoded words that the RFC imposes. It now respects the limit, and
4891 generates multiple encoded words if necessary. To be on the safe side, I
4892 have increased the buffer size for the ${rfc2047: expansion operator from
4895 33. It is now permitted to omit both strings after an "if" condition; if the
4896 condition is true, the result is "true". As before, when the second string
4897 is omitted, a false condition yields an empty string. This makes it less
4898 cumbersome to write custom ACL and router conditions.
4900 34. Failure to deliver a bounce message always caused it to be frozen, even if
4901 there was an errors_to setting on the router. The errors_to setting is now
4904 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4905 canonical form (fully expanded) before being placed in
4906 $sender_host_address.
4908 36. The table in the code that translates DNS record types into text (T_A to
4909 "A" for instance) was missing entries for NS and CNAME. It is just possible
4910 that this could have caused confusion if both these types were looked up
4911 for the same domain, because the text type is used as part of Exim's
4912 per-process caching. But the chance of anyone hitting this buglet seems
4915 37. The dnsdb lookup has been extended in a number of ways.
4917 (1) There is a new type, "zns", which walks up the domain tree until it
4918 finds some nameserver records. It should be used with care.
4920 (2) There is a new type, "mxh", which is like "mx" except that it returns
4921 just the host names, not the priorities.
4923 (3) It is now possible to give a list of domains (or IP addresses) to be
4924 looked up. The behaviour when one of the lookups defers can be
4925 controlled by a keyword.
4927 (4) It is now possible to specify the separator character for use when
4928 multiple records are returned.
4930 38. The dnslists ACL condition has been extended: it is now possible to supply
4931 a list of IP addresses and/or domains to be looked up in a particular DNS
4934 39. Added log_selector=+queue_time_overall.
4936 40. When running the queue in the test harness, wait just a tad after forking a
4937 delivery process, to get repeatability of debugging output.
4939 41. Include certificate and key file names in error message when GnuTLS fails
4940 to set them up, because the GnuTLS error message doesn't include the name
4941 of the failing file when there is a problem reading it.
4943 42. Allow both -bf and -bF in the same test run.
4945 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4947 44. The "Exiscan patch" is now merged into the mainline Exim source.
4949 45. Sometimes the final signoff response after QUIT could fail to get
4950 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4951 < 0 before doing a fflush(). This bug looks as though it goes back to the
4952 introduction of TLS in release 3.20, but "sometimes" must have been rare
4953 because the tests only now provoked it.
4955 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4956 (this can affect the format of dates).
4958 47. exim_tidydb, when checking for the continued existence of a message for
4959 which it has found a message-specific retry record, was not finding
4960 messages that were in split spool directories. Consequently, it was
4961 deleting retry records that should have stayed in existence.
4963 48. Steve fixed some bugs in eximstats.
4965 49. The SPA authentication driver was not abandoning authentication and moving
4966 on to the next authenticator when an expansion was forced to fail,
4967 contradicting the general specification for all authenticators. Instead it
4968 was generating a temporary error. It now behaves as specified.
4970 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4971 (the order specifies the preference for clients). The order is now AES256,
4972 AES128, 3DES, ARCFOUR128.
4974 51. Small patch to Sieve code - explicitly set From: when generating an
4977 52. Exim crashed if a remote delivery caused a very long error message to be
4978 recorded - for instance if somebody sent an entire SpamAssassin report back
4979 as a large number of 550 error lines. This bug was coincidentally fixed by
4980 increasing the size of one of Exim's internal buffers (big_buffer) that
4981 happened as part of the Exiscan merge. However, to be on the safe side, I
4982 have made the code more robust (and fixed the comments that describe what
4985 53. Now that there can be additional text after "Completed" in log lines (if
4986 the queue_time_overall log selector is set), a one-byte patch to exigrep
4987 was needed to allow it to recognize "Completed" as not the last thing in
4990 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4991 patch that reportedly fixes this has been added. I am not expert enough to
4992 create a test for it. This is what the patch creator wrote:
4994 "I found a little strange behaviour of ldap code when working with
4995 Windows 2003 AD Domain, where users was placed in more than one
4996 Organization Units. When I tried to give exim partial DN, the exit code
4997 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4998 But simultaneously result of request was absolutely normal ldap result,
4999 so I produce this patch..."
5001 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5002 REFERENCE, so I have modified the code to exclude the patch when that macro
5005 55. Some experimental protocols are using DNS PTR records for new purposes. The
5006 keys for these records are domain names, not reversed IP addresses. The
5007 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5008 leaves it alone. Component reversal etc. now happens only for IP addresses.
5011 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5013 57. Double the size of the debug message buffer (to 2048) so that more of very
5014 long debug lines gets shown.
5016 58. The exicyclog utility now does better if the number of log files to keep
5017 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5019 59. Two changes related to the smtp_active_hostname option:
5021 (1) $smtp_active_hostname is now available as a variable.
5022 (2) The default for smtp_banner uses $smtp_active_hostname instead
5023 of $primary_hostname.
5025 60. The host_aton() function is supposed to be passed a string that is known
5026 to be a valid IP address. However, in the case of IPv6 addresses, it was
5027 not checking this. This is a hostage to fortune. Exim now panics and dies
5028 if the condition is not met. A case was found where this could be provoked
5029 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5030 components; fortuitously, this particular loophole had already been fixed
5031 by change 4.50/55 above.
5033 If there are any other similar loopholes, the new check in host_aton()
5034 itself should stop them being exploited. The report I received stated that
5035 data on the command line could provoke the exploit when Exim was running as
5036 exim, but did not say which command line option was involved. All I could
5037 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5038 running as the user.
5041 61. There was a buffer overflow vulnerability in the SPA authentication code
5042 (which came originally from the Samba project). I have added a test to the
5043 spa_base64_to_bits() function which I hope fixes it.
5046 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5047 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5049 63. The daemon start-up calls getloadavg() while still root for those OS that
5050 need the first call to be done as root, but it missed one case: when
5051 deliver_queue_load_max is set with deliver_drop_privilege. This is
5052 necessary for the benefit of the queue runner, because there is no re-exec
5053 when deliver_drop_privilege is set.
5055 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5056 This has been fixed.
5058 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5059 was in use, was not putting the data itself into the right store pool;
5060 consequently, it could be overwritten for a subsequent message in the same
5061 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5064 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5066 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5067 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5068 the function were treating the return as a boolean value, which happened to
5069 work because 0=false and not-0=true, but is not correct code.
5071 68. The host_aton() function was not handling scoped IPv6 addresses (those
5072 with, for example, "%eth0" on the end) correctly.
5074 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5075 negated items (that is, ~something) in unsigned ints. Some compilers
5076 apparently mutter when there is no cast.
5078 70. If an address verification called from an ACL failed, and did not produce a
5079 user-specific message (i.e. there was only a "system" message), nothing was
5080 put in $acl_verify_message. In this situation, it now puts the system
5083 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5084 session; change 4.31/43 added the unwanted input to the log line - except
5085 that it did not do this in the start of session case. It now does.
5087 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5088 This is wrong and can cause the other end to generate a synchronization
5089 error if it is another Exim or anything else that does the synchronization
5090 check. A QUIT command is no longer sent after a timeout.
5092 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5093 during host lookups.
5095 74. The defer_ok option of callout verification was not working if it was used
5096 when verifying addresses in header lines, that is, for this case:
5098 verify = header_sender/callout=defer_ok
5100 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5101 those file descriptors could be used for SMTP connections. If anything
5102 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5103 could be sent to the SMTP client, causing chaos. The daemon now opens
5104 stdin, stdout, and stderr to /dev/null when it puts itself into the
5107 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5108 log by default. The user can override this with suitable Perl magic.
5110 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5111 the log message when discard triggers, was not working for the DATA ACL or
5112 for the non-SMTP ACL.
5114 78. Error message wording change in sieve.c.
5116 79. If smtp_accept_max_per_host was set, the number of connections could be
5117 restricted to fewer than expected, because the daemon was trying to set up
5118 a new connection before checking whether the processes handling previous
5119 connections had finished. The check for completed processes is now done
5120 earlier. On busy systems, this bug wouldn't be noticed because something
5121 else would have woken the daemon, and it would have reaped the completed
5124 80. If a message was submitted locally by a user whose login name contained one
5125 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5126 It caused a spool format error. I have fixed the spool reading code. A
5127 related problem was that the "from" clause in the Received: line became
5128 illegal because of the space(s). It is now covered by ${quote_local_part.
5130 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5133 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5135 83. Patch from Sieve maintainer to fix unterminated string problem in
5136 "vacation" handling.
5138 84. Some minor changes to the Linux configuration files to help with other
5139 OS variants using glibc.
5141 85. One more patch for Sieve to update vacation handling to latest spec.
5144 ----------------------------------------------------
5145 See the note above about the 4.44 and 4.50 releases.
5146 ----------------------------------------------------
5152 1. Change 4.43/35 introduced a bug that caused file counts to be
5153 incorrectly computed when quota_filecount was set in an appendfile
5156 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5157 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5160 3. Give more explanation in the error message when the command for a transport
5161 filter fails to execute.
5163 4. There are several places where Exim runs a non-Exim command in a
5164 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5165 was being done only for the command run by the queryprogram router. It is
5166 now done for all such subprocesses. The other cases are: ${run, transport
5167 filters, and the commands run by the lmtp and pipe transports.
5169 5. Some older OS have a limit of 256 on the maximum number of file
5170 descriptors. Exim was using setrlimit() to set 1000 as a large value
5171 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5172 systems. I've change it so that if it can't get 1000, it tries for 256.
5174 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5175 was an oversight, and furthermore, ever since the addition of extra
5176 controls (e.g. 4.43/32), the checks on when to allow different forms of
5177 "control" were broken. There should now be diagnostics for all cases when a
5178 control that does not make sense is encountered.
5180 7. $recipients is now available in the predata ACL (oversight).
5182 8. Tidy the search cache before the fork to do a delivery from a message
5183 received from the command line. Otherwise the child will trigger a lookup
5184 failure and thereby defer the delivery if it tries to use (for example) a
5185 cached ldap connection that the parent has called unbind on.
5187 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5188 of $address_data from the recipient verification was clobbered by the
5189 sender verification.
5191 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5192 was its contents. (It was OK if the option was not defined at all.)
5194 11. A "Completed" log line is now written for messages that are removed from
5195 the spool by the -Mrm option.
5197 12. $host_address is now set to the target address during the checking of
5198 ignore_target_hosts.
5200 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5201 being passed; this would have caused $sender_host_name to have been used if
5202 matching the list had actually called for a host name (not very likely,
5203 since this list is usually IP addresses). A host name is now passed as
5206 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5207 code that specifies a non-restarting handler (typically sigaction() in
5208 modern systems) in an attempt to fix a rare and obscure crash bug.
5210 15. Narrowed the window for a race in the daemon that could cause it to ignore
5211 SIGCHLD signals. This is not a major problem, because they are used only to
5212 wake it up if nothing else does.
5214 16. A malformed maildirsize file could cause Exim to calculate negative values
5215 for the mailbox size or file count. Odd effects could occur as a result.
5216 The maildirsize information is now recalculated if the size or filecount
5219 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5220 support for a long time. Removed HAVE_SYS_VFS_H.
5222 18. Updated exipick to current release from John Jetmore.
5224 19. Allow an empty sender to be matched against a lookup in an address list.
5225 Previously the only cases considered were a regular expression, or an
5228 20. Exim went into a mad DNS lookup loop when doing a callout where the
5229 host was specified on the transport, if the DNS lookup yielded more than
5232 21. The RFC2047 encoding function was originally intended for short strings
5233 such as real names; it was not keeping to the 75-character limit for
5234 encoded words that the RFC imposes. It now respects the limit, and
5235 generates multiple encoded words if necessary. To be on the safe side, I
5236 have increased the buffer size for the ${rfc2047: expansion operator from
5239 22. Failure to deliver a bounce message always caused it to be frozen, even if
5240 there was an errors_to setting on the router. The errors_to setting is now
5243 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5244 canonical form (fully expanded) before being placed in
5245 $sender_host_address.
5247 24. Updated eximstats to version 1.33
5249 25. Include certificate and key file names in error message when GnuTLS fails
5250 to set them up, because the GnuTLS error message doesn't include the name
5251 of the failing file when there is a problem reading it.
5253 26. Expand error message when OpenSSL has problems setting up cert/key files.
5256 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5257 (this can affect the format of dates).
5259 28. exim_tidydb, when checking for the continued existence of a message for
5260 which it has found a message-specific retry record, was not finding
5261 messages that were in split spool directories. Consequently, it was
5262 deleting retry records that should have stayed in existence.
5264 29. eximstats updated to version 1.35
5265 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5266 1.35 - bugfix such that pie charts by volume are generated correctly
5268 30. The SPA authentication driver was not abandoning authentication and moving
5269 on to the next authenticator when an expansion was forced to fail,
5270 contradicting the general specification for all authenticators. Instead it
5271 was generating a temporary error. It now behaves as specified.
5273 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5274 (the order specifies the preference for clients). The order is now AES256,
5275 AES128, 3DES, ARCFOUR128.
5277 31. Small patch to Sieve code - explicitly set From: when generating an
5280 32. Exim crashed if a remote delivery caused a very long error message to be
5281 recorded - for instance if somebody sent an entire SpamAssassin report back
5282 as a large number of 550 error lines. This bug was coincidentally fixed by
5283 increasing the size of one of Exim's internal buffers (big_buffer) that
5284 happened as part of the Exiscan merge. However, to be on the safe side, I
5285 have made the code more robust (and fixed the comments that describe what
5288 33. Some experimental protocols are using DNS PTR records for new purposes. The
5289 keys for these records are domain names, not reversed IP addresses. The
5290 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5291 leaves it alone. Component reversal etc. now happens only for IP addresses.
5294 34. The host_aton() function is supposed to be passed a string that is known
5295 to be a valid IP address. However, in the case of IPv6 addresses, it was
5296 not checking this. This is a hostage to fortune. Exim now panics and dies
5297 if the condition is not met. A case was found where this could be provoked
5298 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5299 components; fortuitously, this particular loophole had already been fixed
5300 by change 4.50/55 or 4.44/33 above.
5302 If there are any other similar loopholes, the new check in host_aton()
5303 itself should stop them being exploited. The report I received stated that
5304 data on the command line could provoke the exploit when Exim was running as
5305 exim, but did not say which command line option was involved. All I could
5306 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5307 running as the user.
5310 35. There was a buffer overflow vulnerability in the SPA authentication code
5311 (which came originally from the Samba project). I have added a test to the
5312 spa_base64_to_bits() function which I hope fixes it.
5315 36. The daemon start-up calls getloadavg() while still root for those OS that
5316 need the first call to be done as root, but it missed one case: when
5317 deliver_queue_load_max is set with deliver_drop_privilege. This is
5318 necessary for the benefit of the queue runner, because there is no re-exec
5319 when deliver_drop_privilege is set.
5321 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5322 was in use, was not putting the data itself into the right store pool;
5323 consequently, it could be overwritten for a subsequent message in the same
5324 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5327 38. Sometimes the final signoff response after QUIT could fail to get
5328 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5329 < 0 before doing a fflush(). This bug looks as though it goes back to the
5330 introduction of TLS in release 3.20, but "sometimes" must have been rare
5331 because the tests only now provoked it.
5337 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5338 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5339 Now it returns an integer. A number of calls were still expecting a T/F
5340 return. Fortuitously, in all cases, the tests worked in OK situations,
5341 which is the norm. However, things would have gone wrong on any write
5342 failures on the smtp file descriptor. This function is used when sending
5343 messages over SMTP and also when doing verify callouts.
5345 2. When Exim is called to do synchronous delivery of a locally submitted
5346 message (the -odf or -odi options), it no longer closes stderr before doing
5349 3. Implemented the mua_wrapper option.
5351 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5353 5. Implemented the functions header_remove(), header_testname(),
5354 header_add_at_position(), and receive_remove_recipient(), and exported them
5357 6. If an ACL "warn" statement specified the addition of headers, Exim already
5358 inserted X-ACL-Warn: at the start if there was no header name. However, it
5359 was not making this test for the second and subsequent header lines if
5360 there were newlines in the string. This meant that an invalid header could
5361 be inserted if Exim was badly configured.
5363 7. Allow an ACL "warn" statement to add header lines at the start or after all
5364 the Received: headers, as well as at the end.
5366 8. Added the rcpt_4xx retry error code.
5368 9. Added postmaster_mailfrom=xxx to callout verification option.
5370 10. Added mailfrom=xxxx to the callout verification option, for verify=
5373 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5374 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5376 12. Inserted some casts to stop certain compilers warning when using pointer
5377 differences as field lengths or precisions in printf-type calls (mostly
5378 affecting debugging statements).
5380 13. Added optional readline() support for -be (dynamically loaded).
5382 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5383 same clock tick as a message's arrival, so that its received time was the
5384 same as the "first fail" time on the retry record, and that message
5385 remained on the queue past the ultimate address timeout, every queue runner
5386 would try a delivery (because it was past the ultimate address timeout) but
5387 after another failure, the ultimate address timeout, which should have then
5388 bounced the address, did not kick in. This was a "< instead of <=" error;
5389 in most cases the first failure would have been in the next clock tick
5390 after the received time, and all would be well.
5392 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5393 being recognized when the domain list was tested by the match_domain
5394 condition in an expansion string.
5396 16. Added the ${str2b64: operator.
5398 17. Exim was always calling setrlimit() to set a large limit for the number of
5399 processes, without checking whether the existing limit was already
5400 adequate. (It did check for the limit on file descriptors.) Furthermore,
5401 errors from getrlimit() and setrlimit() were being ignored. Now they are
5402 logged to the main and panic logs, but Exim does carry on, to try to do its
5403 job under whatever limits there are.
5405 18. Imported PCRE 5.0.
5407 19. Trivial typo in log message " temporarily refused connection" (the leading
5410 20. If the log selector return_path_on_delivery was set and an address was
5411 redirected to /dev/null, the delivery process crashed because it assumed
5412 that a return path would always be set for a "successful" delivery. In this
5413 case, the whole delivery is bypassed as an optimization, and therefore no
5416 21. Internal re-arrangement: the function for sending a challenge and reading
5417 a response while authentication was assuming a zero-terminated challenge
5418 string. It's now changed to take a pointer and a length, to allow for
5419 binary data in such strings.
5421 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5423 23. Exim was not respecting finduser_retries when seeking the login of the
5424 uid under which it was called; it was always trying 10 times. (The default
5425 setting of finduser_retries is zero.) Also, it was sleeping after the final
5426 failure, which is pointless.
5428 24. Implemented tls_on_connect_ports.
5430 25. Implemented acl_smtp_predata.
5432 26. If the domain in control=submission is set empty, Exim assumes that the
5433 authenticated id is a complete email address when it generates From: or
5434 Sender: header lines.
5436 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5437 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5438 chown and chgrp in /bin and hostname in /usr/bin.
5440 28. Exim was keeping the "process log" file open after each use, just as it
5441 does for the main log. This opens the possibility of it remaining open for
5442 long periods when the USR1 signal hits a daemon. Occasional processlog
5443 errors were reported, that could have been caused by this. Anyway, it seems
5444 much more sensible not to leave this file open at all, so that is what now
5447 29. The long-running daemon process does not normally write to the log once it
5448 has entered its main loop, and it closes the log before doing so. This is
5449 so that log files can straightforwardly be renamed and moved. However,
5450 there are a couple of unusual error situations where the daemon does write
5451 log entries, and I had neglected to close the log afterwards.
5453 30. The text of an SMTP error response that was received during a remote
5454 delivery was being truncated at 512 bytes. This is too short for some of
5455 the long messages that one sometimes sees. I've increased the limit to
5458 31. It is now possible to make retry rules that apply only when a message has a
5459 specific sender, in particular, an empty sender.
5461 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5462 it possible to be selective about when SMTP synchronization is enforced.
5464 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5466 32. Implemented hosts_connection_nolog.
5468 33. Added an ACL for QUIT.
5470 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5473 35. Added mailbox_size and mailbox_filecount to appendfile.
5475 36. Added control = no_multiline_responses to ACLs.
5477 37. There was a bug in the logic of the code that waits for the clock to tick
5478 in the case where the clock went backwards by a substantial amount such
5479 that the microsecond fraction of "now" was more than the microsecond
5480 fraction of "then" (but the whole seconds number was less).
5482 38. Added support for the libradius Radius client library this is found on
5483 FreeBSD (previously only the radiusclient library was supported).
5489 1. When certain lookups returned multiple values in the form name=value, the
5490 quoting of the values was not always being done properly. Specifically:
5491 (a) If the value started with a double quote, but contained no whitespace,
5493 (b) If the value contained whitespace other than a space character (i.e.
5494 tabs or newlines or carriage returns) it was not quoted.
5495 This fix has been applied to the mysql and pgsql lookups by writing a
5496 separate quoting function and calling it from the lookup code. The fix
5497 should probably also be applied to nisplus, ibase and oracle lookups, but
5498 since I cannot test any of those, I have not disturbed their existing code.
5500 2. A hit in the callout cache for a specific address caused a log line with no
5501 reason for rejecting RCPT. Now it says "Previous (cached) callout
5502 verification failure".
5504 3. There was an off-by-one bug in the queryprogram router. An over-long
5505 return line was truncated at 256 instead of 255 characters, thereby
5506 overflowing its buffer with the terminating zero. As well as fixing this, I
5507 have increased the buffer size to 1024 (and made a note to document this).
5509 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5510 when Exim is waiting for an SMTP response from a remote server, Exim
5511 restarts its select() call on the socket, thereby resetting its timeout.
5512 This is not a problem when such interrupts are rare. Somebody set up a cron
5513 job to run exiwhat every 2 minutes, which is less than the normal select()
5514 timeout (5 or 10 minutes). This meant that the select() timeout never
5515 kicked in because it was always reset. I have fixed this by comparing the
5516 time when an interrupt arrives with the time at the start of the first call
5517 to select(). If more time than the timeout has elapsed, the interrupt is
5518 treated as a timeout.
5520 5. Some internal re-factoring in preparation for the addition of Sieve
5521 extensions (by MH). In particular, the "personal" test is moved to a
5522 separate function, and given an option for scanning Cc: and Bcc: (which is
5523 not set for Exim filters).
5525 6. When Exim created an email address using the login of the caller as the
5526 local part (e.g. when creating a From: or Sender: header line), it was not
5527 quoting the local part when it contained special characters such as @.
5529 7. Installed new OpenBSD configuration files.
5531 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5532 try to make them clearer.
5534 9. Callout options, other than the timeout value, were being ignored when
5535 verifying sender addresses in header lines. For example, when using
5537 verify = header_sender/callout=no_cache
5539 the cache was (incorrectly) being used.
5541 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5542 only the Cygwin environment.
5544 11. When return_path_on_delivery was set as a log selector, if different remote
5545 addresses in the same message used different return paths and parallel
5546 remote delivery occurred, the wrong values would sometimes be logged.
5547 (Whenever a remote delivery process finished, the return path value from
5548 the most recently started remote delivery process was logged.)
5550 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5551 lines when AUTH and/or TLS are in use. This is the "received protocol"
5552 field. Exim used to use "asmtp" for authenticated SMTP, without any
5553 indication (in the protocol name) for TLS use. Now it follows the RFC and
5554 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5555 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5556 names appear in log lines as well as in Received: header lines.
5558 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5559 extensions, and comparison tests, and to fix some bugs.
5561 14. Changes to the "personal" filter test:
5563 (1) The test was buggy in that it was just doing the equivalent of
5564 "contains" tests on header lines. For example, if a user's address was
5565 anne@some.where, the "personal" test would incorrectly be true for
5567 To: susanne@some.where
5569 This test is now done by extracting each address from the header in turn,
5570 and checking the entire address. Other tests that are part of "personal"
5571 are now done using regular expressions (for example, to check local parts
5572 of addresses in From: header lines).
5574 (2) The list of non-personal local parts in From: addresses has been
5575 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5576 taken from the Sieve specification recommendations.
5578 (3) If the message contains any header line starting with "List-" it is
5579 treated as non-personal.
5581 (4) The test for "circular" in the Subject: header line has been removed
5582 because it now seems ill-conceived.
5584 15. Minor typos in src/EDITME comments corrected.
5586 16. Installed latest exipick from John Jetmore.
5588 17. If headers_add on a router specified a text string that was too long for
5589 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5590 of string_sprintf() is now avoided.
5592 18. $message_body_size was not set (it was always zero) when running the DATA
5593 ACL and the local_scan() function.
5595 19. For the "mail" command in an Exim filter, no default was being set for
5596 the once_repeat time, causing a random time value to be used if "once" was
5597 specified. (If the value happened to be <= 0, no repeat happened.) The
5598 default is now 0s, meaning "never repeat". The "vacation" command was OK
5599 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5600 (I found it when inspecting the code).
5602 20. There is now an overall timeout for performing a callout verification. It
5603 defaults to 4 times the callout timeout, which applies to individual SMTP
5604 commands during the callout. The overall timeout applies when there is more
5605 than one host that can be tried. The timeout is checked before trying the
5606 next host. This prevents very long delays if there are a large number of
5607 hosts and all are timing out (e.g. when the network connections are timing
5608 out). The value of the overall timeout can be changed by specifying an
5609 additional sub-option for "callout", called "maxwait". For example:
5611 verify = sender/callout=5s,maxwait=20s
5613 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5614 to the end before writing, but this should make it even safer).
5616 22. Exim was forgetting that it had advertised PIPELINING for the second and
5617 subsequent messages on an SMTP connection. It was also not resetting its
5618 memory on STARTTLS and an internal HELO.
5620 23. When Exim logs an SMTP synchronization error within a session, it now
5621 records whether PIPELINING has been advertised or not.
5623 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5624 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5625 rather than long int.
5627 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5629 26. Added the never_mail option to autoreply.
5635 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5636 crash if the getsockname() call failed; this can happen if a connection is
5637 closed very soon after it is established. The problem was simply in the
5638 order in which certain operations were done, causing Exim to try to write
5639 to the SMTP stream before it had set up the file descriptor. The bug has
5640 been fixed by making things happen in the correct order.
5646 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5647 before the connection was closed, thus losing the rejection response.
5649 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5650 some early Solaris releases, but causes trouble in current releases where
5651 socklen_t is defined.
5653 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5656 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5659 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5660 that releasing the top of it at the end releases what was used for sub-
5661 expansions (unless the block got too big). However, discard this block if
5662 the first thing is a variable or header, so that we can use its block when
5663 it is dynamic (useful for very large $message_headers, for example).
5665 6. Lookups now cache *every* query, not just the most recent. A new, separate
5666 store pool is used for this. It can be recovered when all lookup caches are
5667 flushed. Lookups now release memory at the end of their result strings.
5668 This has involved some general refactoring of the lookup sources.
5670 7. Some code has been added to the store_xxx() functions to reduce the amount
5671 of flapping under certain conditions.
5673 8. log_incoming_interface used to affect only the <= reception log lines. Now
5674 it causes the local interface and port to be added to several more SMTP log
5675 lines, for example "SMTP connection from", and rejection lines.
5677 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5679 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5681 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5682 could be overwritten at the end of the current message (or the start of a
5683 new message if it was set in a HELO ACL). The value is now preserved for
5684 the duration of the SMTP connection.
5686 12. If a transport had a headers_rewrite setting, and a matching header line
5687 contained an unqualified address, that address was qualified, even if it
5688 did not match any rewriting rules. The underlying bug was that the values
5689 of the flags that permit the existence of unqualified sender and recipient
5690 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5691 non-local messages, and by -bnq for local messages) were not being
5692 preserved with the message after it was received.
5694 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5695 "next input=" as part of the text comprising the host identity instead of
5696 the correct text. The code was using the same buffer for two different
5697 strings. However, depending on which order the printing function evaluated
5698 its arguments, the bug did not always show up. Under Linux, for example, my
5699 test suite worked just fine.
5701 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5702 doesn't work with some older versions of Perl. It has been changed to "my",
5703 which in any case is probably the better facility to use.
5705 15. A really picky compiler found some instances of statements for creating
5706 error messages that either had too many or two few arguments for the format
5709 16. The size of the buffer for calls to the DNS resolver has been increased
5710 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5711 for addresses that have a lot of PTR records. This alleviates a problem; it
5712 does not fully solve it.
5714 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5715 buffer now truncates the list and logs the incident, which is the same
5716 action as happens when Exim is looking up a host name and its aliases.
5717 Previously in this situation something unpredictable would happen;
5718 sometimes it was "internal error: store_reset failed".
5720 18. If a server dropped the connection unexpectedly when an Exim client was
5721 using GnuTLS and trying to read a response, the client delivery process
5722 crashed while trying to generate an error log message.
5724 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5725 string, for example:
5727 warn message = H1: something\nH2: something
5729 the text was added as a single header line from Exim's point of view
5730 though it ended up OK in the delivered message. However, searching for the
5731 second and subsequent header lines using $h_h2: did not work. This has been
5732 fixed. Similarly, if a system filter added multiple headers in this way,
5733 the routers could not see them.
5735 20. Expanded the error message when iplsearch is called with an invalid key to
5736 suggest using net-iplsearch in a host list.
5738 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5739 ACLs are no longer actually imposed (and a message to that effect is
5742 22. If a "gecos" field in a passwd entry contained escaped characters, in
5743 particular, if it contained a \" sequence, Exim got it wrong when building
5744 a From: or a Sender: header from that name. A second bug also caused
5745 incorrect handling when an unquoted " was present following a character
5746 that needed quoting.
5748 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5749 was not being matched caselessly.
5751 24. Arranged for all hyphens in the exim.8 source to be escaped with
5754 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5755 properly. Recipient callout cache records were still being keyed to include
5756 the sender, even when use_sender was set false. This led to far more
5757 callouts that were necessary. The sender is no longer included in the key
5758 when use_sender is false.
5760 26. Added "control = submission" modifier to ACLs.
5762 27. Added the ${base62d: operator to decode base 62 numbers.
5764 28. dnsdb lookups can now access SRV records.
5766 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5767 the configuration file.
5769 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5770 (-v) mode. This makes the output for a verbose queue run more intelligible.
5772 31. Added a use_postmaster feature to recipient callouts.
5774 32. Added the $body_zerocount variable, containing the number of binary zero
5775 bytes in the message body.
5777 33. The time of last modification of the "new" subdirectory is now used as the
5778 "mailbox time last read" when there is a quota error for a maildir
5781 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5783 35. Added +ignore_unknown as a special item in host lists.
5785 36. Code for decoding IPv6 addresses in host lists is now included, even if
5786 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5787 address was recognized as an IP address, but was then not correctly decoded
5788 into binary, causing unexpected and incorrect effects when compared with
5795 1. Very minor rewording of debugging text in manualroute to say "list of
5796 hosts" instead of "hostlist".
5798 2. If verify=header_syntax was set, and a header line with an unqualified
5799 address (no domain) and a large number of spaces between the end of the
5800 name and the colon was received, the reception process suffered a buffer
5801 overflow, and (when I tested it) crashed. This was caused by some obsolete
5802 code that should have been removed. The fix is to remove it!
5804 3. When running in the test harness, delay a bit after writing a bounce
5805 message to get a bit more predictability in the log output.
5807 4. Added a call to search_tidyup() just before forking a reception process. In
5808 theory, someone could use a lookup in the expansion of smtp_accept_max_
5809 per_host which, without the tidyup, could leave open a database connection.
5811 5. Added the variables $recipient_data and $sender_data which get set from a
5812 lookup success in an ACL "recipients" or "senders" condition, or a router
5813 "senders" option, similar to $domain_data and $local_part_data.
5815 6. Moved the writing of debug_print from before to after the "senders" test
5818 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5819 problems for message scanning, either using a data ACL, or using
5820 local_scan() because the Received: header was not generated till after they
5821 were called (in order to set the time as the time of reception completion).
5822 I have revised the way this works. The header is now generated after the
5823 body is received, but before the ACL or local_scan() are called. After they
5824 are run, the timestamp in the header is updated.
5830 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5831 before starting a queue runner without re-exec. This happened only when
5832 deliver_drop_privilege was set or when the Exim user was set to root. The
5833 effect of the bug was that timeouts during subsequent deliveries caused
5834 crashes instead of being properly handled. The handler is now left at its
5835 default (and expected) setting.
5837 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5838 message, again when deliver_drop_privilege is set or Exim is run as root.
5839 The bug described in (1) was not present in this case, but the tidying up
5840 of the other signals was missing. I have made the two cases consistent.
5842 3. The ignore_target_hosts setting on a manualroute router was being ignored
5843 for hosts that were looked up using the /MX notation.
5845 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5848 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5849 operated on the sender address. After changing the $sender_address to <>
5850 for the sender address verify, Exim was re-instated it as the original
5851 (before rewriting) address, but remembering that it had rewritten it, so it
5852 wasn't rewriting it again. This bug also had the effect of breaking the
5853 sender address verification caching when the sender address was rewritten.
5855 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5856 This has been changed so that if the ip literal address matches
5857 ignore_target_hosts, the router declines.
5859 7. Added expansion conditions match_domain, match_address, and match_local_
5860 part (NOT match_host).
5862 8. The placeholder for the Received: header didn't have a length field set.
5864 9. Added code to Exim itself and to exim_lock to test for a specific race
5865 condition that could lead to file corruption when using MBX delivery. The
5866 issue is with the lockfile that is created in /tmp. If this file is removed
5867 after a process has opened it but before that process has acquired a lock,
5868 there is the potential for a second process to recreate the file and also
5869 acquire a lock. This could lead to two Exim processes writing to the file
5870 at the same time. The added code performs the same test as UW imapd; it
5871 checks after acquiring the lock that its file descriptor still refers to
5872 the same named file.
5874 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5875 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5878 11. Added the smtp_active_hostname option. If used, this will typically be made
5879 to depend on the incoming interface address. Because $interface_address is
5880 not set up until the daemon has forked a reception process, error responses
5881 that can happen earlier (such as "too many connections") no longer contain
5884 12. If an expansion in a condition on a "warn" statement fails because a lookup
5885 defers, the "warn" statement is abandoned, and the next ACL statement is
5886 processed. Previously this caused the whole ACL to be aborted.
5888 13. Added the iplsearch lookup type.
5890 14. Added ident_timeout as a log selector.
5892 15. Added tls_certificate_verified as a log selector.
5894 16. Added a global option tls_require_ciphers (compare the smtp transport
5895 option of the same name). This controls incoming TLS connections.
5897 17. I finally figured out how to make tls_require_ciphers do a similar thing
5898 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5899 before starting the TLS session.
5901 18. Tabs are now shown as \t in -bP output.
5903 19. If the log selector return_path_on_delivery was set, Exim crashed when
5904 bouncing a message because it had too many Received: header lines.
5906 20. If two routers both had headers_remove settings, and the first one included
5907 a superfluous trailing colon, the final name in the first list and the
5908 first name in the second list were incorrectly joined into one item (with a
5909 colon in the middle).
5915 1. Added -C and -D options to the exinext utility, mainly to make it easier
5916 to include in the automated testing, but these could be helpful when
5917 multiple configurations are in use.
5919 2. The exinext utility was not formatting the output nicely when there was
5920 an alternate port involved in the retry record key, nor when there was a
5921 message id as well (for retries that were specific to a specific message
5922 and a specific host). It was also confused by IPv6 addresses, because of
5923 the additional colons they contain. I have fixed the IPv4 problem, and
5924 patched it up to do a reasonable job for IPv6.
5926 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5927 delivery, the log line now contains "pipelined" if PIPELINING was used.
5929 4. An SMTP transport process used to panic and die if the bind() call to set
5930 an explicit outgoing interface failed. This has been changed; it is now
5931 treated in the same way as a connect() failure.
5933 5. A reference to $sender_host_name in the part of a conditional expansion
5934 that was being skipped was still causing a DNS lookup. This no longer
5937 6. The def: expansion condition was not recognizing references to header lines
5938 that used bh_ and bheader_.
5940 7. Added the _cache feature to named lists.
5942 8. The code for checking quota_filecount in the appendfile transport was
5943 allowing one more file than it should have been.
5945 9. For compatibility with Sendmail, the command line option
5953 and sets the incoming protocol and host name (for trusted callers). The
5954 host name and its colon can be omitted when only the protocol is to be set.
5955 Note the Exim already has two private options, -pd and -ps, that refer to
5956 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5957 "s", but I don't think that's a major issue.
5959 10. A number of refactoring changes to the code, none of which should affect
5962 (a) The number of logging options was getting close to filling up the
5963 32-bit word that was used as a bit map. I have split them into two classes:
5964 those that are passed in the argument to log_write(), and those that are
5965 only ever tested independently outside of that function. These are now in
5966 separate 32-bit words, so there is plenty of room for expansion again.
5967 There is no change in the user interface or the logging behaviour.
5969 (b) When building, for example, log lines, the code previously used a
5970 macro that called string_cat() twice, in order to add two strings. This is
5971 not really sufficiently general. Furthermore, there was one instance where
5972 it was actually wrong because one of the argument was used twice, and in
5973 one call a function was used. (As it happened, calling the function twice
5974 did not affect the overall behaviour.) The macro has been replaced by a
5975 function that can join an arbitrary number of extra strings onto a growing
5978 (c) The code for expansion conditions now uses a table and a binary chop
5979 instead of a serial search (which was left over from when there were very
5980 few conditions). Also, it now recognizes conditions like "pam" even when
5981 the relevant support is not compiled in: a suitably worded error message is
5982 given if an attempt is made to use such a condition.
5984 11. Added ${time_interval:xxxxx}.
5986 12. A bug was causing one of the ddress fields not to be passed back correctly
5987 from remote delivery subprocesses. The field in question was not being
5988 subsequently used, so this caused to problems in practice.
5990 13. Added new log selectors queue_time and deliver_time.
5992 14. Might have fixed a bug in maildirsizefile handling that threw up
5993 "unexpected character" debug warnings, and recalculated the data
5994 unnecessarily. In any case, I expanded the warning message to give more
5997 15. Added the message "Restricted characters in address" to the statements in
5998 the default ACL that block characters like @ and % in local parts.
6000 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6001 Three changes have been made:
6003 (a) There was a serious bug; a negative response to MAIL caused the whole
6004 recipient domain to be cached as invalid, thereby blocking all messages
6005 to all local parts at the same domain, from all senders. This bug has
6006 been fixed. The domain is no longer cached after a negative response to
6007 MAIL if the sender used is not empty.
6009 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6012 (c) A new callout option, "use_sender" has been added for people who want
6013 the modified behaviour.
6019 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6022 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6023 indeed breaks things for older releases.
6025 3. Added additional logging to the case where there is a problem reading data
6026 from a filter that is running in a subprocess using a pipe, in order to
6027 try to track down a specific problem.
6029 4. Testing facility fudge: when running in the test harness and attempting
6030 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6031 getting "No route to host". Convert this to a timeout.
6033 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6036 6. Some OS don't have socklen_t but use size_t instead. This affects the
6037 fifth argument of getsockopt() amongst other things. This is now
6038 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6039 can be set for individual OS. I have set it for SunOS5, OSF1, and
6040 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6041 some earlier ones do not.
6043 7. Change 4.30/15 was not doing the test caselessly.
6045 8. The standard form for an IPv6 address literal was being rejected by address
6046 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6047 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6048 this, as well as the form without the "IPv6" on the front (but only when
6049 address literals are enabled, of course).
6051 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6053 10. Exim crashed if a message with an empty sender address specified by -f
6054 encountered a router with an errors_to setting. This could be provoked only
6055 by a command such as
6059 where an empty string was supplied; "<>" did not hit this bug.
6061 11. Installed PCRE release 4.5.
6063 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6064 remained set. It is now erased.
6066 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6067 times from message ids (which are base 36 rather than the normal 62).
6069 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6070 were being counted as actual protocol errors, and logged if the log
6071 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6072 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6073 and DATA following a set of rejected RCPTs do not count as protocol errors.
6074 In other words, Exim assumes they were pipelined, though this may not
6075 actually be the case. Of course, in all cases the client gets an
6076 appropriate error code.
6078 15. If a lookup fails in an ACL condition, a message about the failure may
6079 be available; it is used if testing the ACL cannot continue, because most
6080 such messages specify what the cause of the deferral is. However, some
6081 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6082 that caused an old message to be retained and used if a later statement
6083 caused a defer, replacing the real cause of the deferral.
6085 16. If an IP address had so many PTR records that the DNS lookup buffer
6086 was not large enough to hold them, Exim could crash while trying to process
6087 the truncated data. It now detects and logs this case.
6089 17. Further to 4.21/58, another change has been made: if (and only if) the
6090 first line of a message (the first header line) ends with CRLF, a bare LF
6091 in a subsequent header line has a space inserted after it, so as not to
6092 terminate the header.
6094 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6095 unnecessarily, used atoi() instead of strtol(), and didn't check the
6096 termination when getting file sizes from file names by regex.
6098 19. Completely re-implemented the support for maildirsize files, in the light
6099 of a number of problems with the previous contributed implementation
6100 (4.30/29). In particular:
6102 . If the quota is zero, the maildirsize file is maintained, but no quota is
6105 . If the maildir directory does not exist, it is created before any attempt
6106 to write a maildirsize file.
6108 . The quota value in the file is just a cache; if the quota is changed in
6109 the transport, the new value overrides.
6111 . A regular expression is available for excluding directories from the
6114 20. The autoreply transport checks the characters in options that define the
6115 message's headers; it allows continued headers, but it was checking with
6116 isspace() after an embedded newline instead of explicitly looking for a
6119 21. If all the "regular" hosts to which an address was routed had passed their
6120 expiry times, and had not reached their retry times, the address was
6121 bounced, even if fallback hosts were defined. Now Exim should go on to try
6124 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6125 equivalent code in the SMTP transport. Some hosts send humungous responses
6126 to HELO/EHLO, more than 1024 it seems.
6128 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6129 gives compiler warnings in some environments. I've now done it "properly",
6132 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6133 (because of problems with the built-in one) was declared to return uschar *
6134 instead of char *, causing compiler failure.
6136 25. Fixed a file descriptor leak when processing alias/forward files.
6138 26. Fixed a minor format string issue in dbfn.c.
6140 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6142 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6143 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6146 29. When a sender address is verified, it is cached, to save repeating the test
6147 when there is more than one recipient in a message. However, when the
6148 verification involves a callout, it is possible for different callout
6149 options to be set for different recipients. It is too complicated to keep
6150 track of this in the cache, so now Exim always runs a verification when a
6151 callout is required, relying on the callout cache for the optimization.
6152 The overhead is duplication of the address routing, but this should not be
6155 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6156 to be verified with callout=postmaster, and the main callout worked but the
6157 postmaster check failed, the verification correctly failed. However, if a
6158 subsequent RCPT command asked for sender verification *without* the
6159 postmaster check, incorrect caching caused this verification also to fail,
6162 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6163 it was not caching the DNS options (qualify_single, search_parents) that
6164 were used when the lookup failed. A subsequent lookup with different
6165 options therefore always gave the same answer, though there were cases
6166 where it should not have. (Example: a "domains = !$mx_any" option on a
6167 dnslookup router: the "domains" option is always processed without any
6168 widening, but the router might have qualify_single set.) Now Exim uses the
6169 cached value only when the same options are set.
6171 32. Added John Jetmore's "exipick" utility to the distribution.
6173 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6174 than a timeout (e.g. a certificate is required, and is not provided), an
6175 Exim server now closes the connection immediately. Previously it waited for
6176 the client to close - but if the client is SSL, it seems that they each
6177 wait for each other, leading to a delay before one of them times out.
6179 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6180 maintained 0.8.x compatibility because I don't think many are using it, and
6181 it is clearly obsolete.
6183 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6186 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6187 client certificate was expired. A simple patch fixes this, though I don't
6188 understand the full logic of why the verify callback is called multiple
6191 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6192 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6193 which causes problems with some clients (such as the Certicom SSL Plus
6194 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6195 disables the coutermeasure allowing Eudora to connect."
6197 38. Exim was not checking that a write() to a log file succeeded. This could
6198 lead to Bad Things if a log got too big, in particular if it hit a file
6199 size limit. Exim now panics and dies if it cannot write to a log file, just
6200 as it does if it cannot open a log file.
6202 39. Modified OS/Makefile-Linux so that it now contains
6204 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6206 The two -D definitions ensure that Exim is compiled with large file
6207 support, which makes it possible to handle log files that are bigger than
6210 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6211 instance) a domain was checked against a named list that involved a lookup,
6212 causing $domain_data to be set, then another domain was checked against the
6213 same list, then the first domain was re-checked, the value of $domain_data
6214 after the final check could be wrong. In particular, if the second check
6215 failed, it could be set empty. This bug probably also applied to
6218 41. The strip_trailing_dot option was not being applied to the address given
6219 with the -f command-line option.
6221 42. The code for reading a message's header from the spool was incrementing
6222 $received_count, but never initializing it. This meant that the value was
6223 incorrect (doubled) while delivering a message in the same process in which
6224 it was received. In the most common configuration of Exim, this never
6225 happens - a fresh exec is done - but it can happen when
6226 deliver_drop_privilege is set.
6228 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6229 it now includes up to 150 characters of the unexpected data in the log
6232 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6233 and building data strings. The size of both of these buffers was 10 000
6234 bytes - far larger than anybody would *ever* want, thought I. Needless to
6235 say, somebody hit the limit. I have increased the maximum line length to
6236 20 000 and the maximum data length of concatenated lines to 100 000. I have
6237 also fixed two bugs, because there was no checking on these buffers. Tsk,
6238 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6239 buffer is too small.
6241 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6242 lsearch lookups. Now it does.
6244 46. When parsing a route_list item in a manualroute router, a fixed-length
6245 buffer was used for the list of hosts. I made this 1024 bytes long,
6246 thinking that nobody would ever have a list of hosts that long. Wrong.
6247 Somebody had a whole pile of complicated expansion conditions, and the
6248 string was silently truncated, leading to an expansion error. It turns out
6249 that it is easier to change to an unlimited length (owing to other changes
6250 that have happened since this code was originally written) than to build
6251 structure for giving a limitation error. The length of the item that
6252 expands into the list of hosts is now unlimited.
6254 47. The lsearch lookup could not handle data where the length of text line was
6255 more than 4095 characters. Such lines were truncated, leading to shortened
6256 data being returned. It should now handle lines of any length.
6258 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6259 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6262 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6263 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6264 no longer happens in comment lines. A list of replacements is now placed
6265 at the head of all of the source files, except those whose only change is
6266 to replace PERL_COMMAND in the very first #! line.
6268 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6269 messages on the queue, with a bottom-up merge sort, using code contributed
6270 by Michael Haardt. This should make operations like -bp somewhat faster on
6271 large queues. It won't affect queue runners, except when queue_run_in_order
6274 51. Installed eximstats 1.31 in the distribution.
6276 52. Added support for SRV lookups to the dnslookup router.
6278 53. If an ACL referred to $message_body or $message_body_end, the value was not
6279 reset for any messages that followed in the same SMTP session.
6281 54. The store-handling optimization for building very long strings was not
6282 differentiating between the different store pools. I don't think this
6283 actually made any difference in practice, but I've tidied it.
6285 55. While running the routers to verify a sender address, $sender_address
6286 was still set to the sender address. This is wrong, because when routing to
6287 send a bounce to the sender, it would be empty. Therefore, I have changed
6288 it so that, while verifying a sender address, $sender_address is set to <>.
6289 (There is no change to what happens when verifying a recipient address.)
6291 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6292 target A or AAAA records (if not already returned) without resetting the
6293 qualify_single or search_parents options of the DNS resolver. These are
6294 inappropriate in this case because the targets of MX and SRV records must
6295 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6296 target that, when qualified, matched something in the local domain. These
6297 two options are now turned off when doing these lookups.
6299 57. It seems that at least some releases of Reiserfs (which does not have the
6300 concept of a fixed number of inodes) returns zero and not -1 for the
6301 number of available inodes. This interacted badly with check_spool_inodes,
6302 which assumed that -1 was the "no such thing" setting. What I have done is
6303 to check that the total number of inodes is greater than zero before doing
6304 the test of how many are available.
6306 58. When a "warn" ACL statement has a log_message modifier, the message is
6307 remembered, and not repeated. This is to avoid a lot of repetition when a
6308 message has many recipients that cause the same warning to be written.
6309 However, Exim was preserving the list of already written lines for an
6310 entire SMTP session, which doesn't seem right. The memory is now reset if a
6311 new message is started.
6313 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6314 debugging output unless log_rewrite was also set.
6316 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6317 of (int)(handle) when we know that handle contains (void *)(-1).
6319 61. The Exim daemon panic-logs an error return when it closes the incoming
6320 connection. However "connection reset by peer" seems to be common, and
6321 isn't really an error worthy of noting specially, so that particular error
6324 62. When Exim is trying to find all the local interfaces, it used to panic and
6325 die if the ioctl to get the interface flags failed. However, it seems that
6326 on at least one OS (Solaris 9) it is possible to have an interface that is
6327 included in the list of interfaces, but for which you get a failure error
6328 for this call. This happens when the interface is not "plumbed" into a
6329 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6330 failure of the "get flags" call assumes that the interface is down.
6332 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6333 makes life easier for people who are doing arithmetic on fields extracted
6334 from dates, where you often get leading zeros that should not be
6335 interpreted as octal.
6337 64. Added qualify_domain to the redirect router, to override the global
6340 65. If a pathologically long header line contained very many addresses (the
6341 report of this problem mentioned 10 000) and each of them was rewritten,
6342 Exim could use up a very large amount of memory. (It kept on making new
6343 copies of the header line as it rewrote, and never released the old ones.)
6344 At the expense of a bit more processing, the header rewriting function has
6345 been changed so that it no longer eats memory in this way.
6347 66. The generation of the Received: header has been moved from the time that a
6348 message starts to be received, to the time that it finishes. The timestamp
6349 in the Received: header should now be very close to that of the <= log
6350 line. There are two side-effects of this change:
6352 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6353 logged header lines no longer include the local Received: line, because
6354 it has not yet been created. The same applies to a copy of the message
6355 that is returned to a non-SMTP sender when a message is rejected.
6357 (b) When a filter file is tested using -bf, no additional Received: header
6358 is added to the test message. After some thought, I decided that this
6361 This change does not affect the value of $received_for. It is still set
6362 after address rewriting, but before local_scan() is called.
6364 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6366 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6367 gave an unhelpful panic error message, and a defer error. I have managed to
6368 change this behaviour so that it now rejects any supplied certificate,
6369 which seems right, as the list of acceptable certificates is empty.
6371 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6372 gave an unhelpful defer error. I have not managed to make this reject any
6373 supplied certificates, but the error message it gives is "no certificate
6374 supplied", which is not helpful.
6376 70. exigrep's output now also includes lines that are not associated with any
6377 message, but which match the given pattern. Implemented by a patch from
6378 Martin Sluka, which also tidied up the Perl a bit.
6380 71. Recipient callout verification, like sender verification, was using <> in
6381 the MAIL FROM command. This isn't really the right thing, since the actual
6382 sender may affect whether the remote host accepts the recipient or not. I
6383 have changed it to use the actual sender in the callout; this means that
6384 the cache record is now keyed on a recipient/sender pair, not just the
6385 recipient address. There doesn't seem to be a real danger of callout loops,
6386 since a callout by the remote host to check the sender would use <>.
6387 [SEE ABOVE: changed after hitting problems.]
6389 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6390 temporary errors. However, in the case of such a code being given after
6391 the end of a data transmission (i.e. after ".") Exim was failing to write
6392 a retry record for the message. (Yes, there was some broken host that was
6393 actually sending 8xx at this point.)
6395 73. An unknown lookup type in a host list could cause Exim to panic-die when
6396 the list was checked. (An example that provoked this was putting <; in the
6397 middle of a list instead of at the start.) If this happened during a DATA
6398 ACL check, a -D file could be left lying around. This kind of configuration
6399 error no longer causes Exim to die; instead it causes a defer error. The
6400 incident is still logged to the main and panic logs.
6402 74. Buglet left over from Exim 3 conversion. The message "too many messages
6403 in one connection" was written to the rejectlog but not the mainlog, except
6404 when address rewriting (yes!) was being logged.
6406 75. Added write_rejectlog option.
6408 76. When a system filter was run not as root (that is, when system_filter_user
6409 was set), the values of the $n variables were not being returned to the
6410 main process; thus, they were not subsequently available in the $sn
6413 77. Added +return_path_on_delivery log selector.
6415 78. A connection timeout was being treated differently from recipients deferred
6416 when testing hosts_max_try with a message that was older than the host's
6417 retry timeout. (The host should not be counted, thus allowing all hosts to
6418 be tried at least once before bouncing.) This may have been the cause of an
6419 occasionally reported bug whereby a message would remain on the queue
6420 longer than the retry timeout, but would be bounced if a delivery was
6421 forced. I say "may" because I never totally pinned down the problem;
6422 setting up timeout/retry tests is difficult. See also the next item.
6424 79. The ultimate address timeout was not being applied to errors that involved
6425 a combination of host plus message (for example, a timeout on a MAIL
6426 command). When an address resolved to a number of possible hosts, and they
6427 were not all tried for each delivery (e.g. because of hosts_max_try), a
6428 message could remain on the queue longer than the retry timeout.
6430 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6433 81. Fixed an obscure SMTP outgoing bug which required at least the following
6434 conditions: (a) there was another message waiting for the same server;
6435 (b) the server returned 5xx to all RCPT commands in the first message so
6436 that the message was not completed; (c) the server dropped the connection
6437 or gave a negative response to the RSET that Exim sends to abort the
6438 transaction. The observed case was a dropped connection after DATA that had
6439 been sent in pipelining mode. That is, the server had advertised PIPELINING
6440 but was not implementing it correctly. The effect of the bug was incorrect
6441 behaviour, such as trying another host, and this could lead to a crash.
6447 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6448 and daemon.c were passed as pointers to ints; they should have been
6449 pointers to socklen_t variables (which are typically unsigned ints).
6451 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6454 3. Fixed a really odd bug that affected only the testing scheme; patching a
6455 certain fixed string in the binary changed the value of another string that
6456 happened to be identical to the end of the original first string.
6458 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6459 name", it returns that address as the IP address. On some operating
6460 systems (e.g. Solaris), it also passes back the IP address string as the
6461 "host name". However, on others (e.g. Linux), it passes back an empty
6462 string. Exim wasn't checking for this, and was changing the host name to an
6463 empty string, assuming it had been canonicalized.
6465 5. Although rare, it is permitted to have more than one PTR record for a given
6466 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6467 all the names associated with an address, because they do in Solaris.
6468 However, it seems that they do not in Linux for data that comes from the
6469 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6470 I found this out when I moved to a new Linux workstation and tried to run
6471 the Exim test suite.
6473 To get round this problem I have changed the code so that it now does its
6474 own call to the DNS to look up PTR records when searching for a host name.
6475 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6476 addresses that are only in /etc/hosts are still found.
6478 This behaviour is, however, controlled by an option called host_lookup_
6479 order, which defaults to "bydns:byaddr". If people want to use the other
6480 order, or indeed, just use one or the other means of lookup, they can
6481 specify it in this variable.
6483 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6484 some operating systems, this comes back from gethostbyaddr() as an empty
6485 string, and this is what Exim used to test for. However, it seems that in
6486 other systems, "." is yielded. Exim now tests for this case too.
6488 7. The values of check_spool_space and check_log_space are now held internally
6489 as a number of kilobytes instead of an absolute number of bytes. If a
6490 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6491 kilobyte. This means that much larger values can be stored.
6493 8. Exim monitor: an attempt to get the action menu when not actually pointing
6494 at a message produces an empty menu entitled "No message selected". This
6495 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6496 no entries in it ("Shell widget menu has zero width and/or height"). So I
6497 have added a single, blank menu entry in this case.
6499 9. Added ${quote_local_part.
6501 10. MIME decoding is now applied to the contents of Subject: header lines when
6504 11. Now that a reference to $sender_host_address automatically causes a reverse
6505 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6506 host lookup before query-style lookups in lists that might use this
6507 variable. This has therefore been abolished, and the "net-" prefix is no
6508 longer necessary for query-style lookups.
6510 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6511 have been a typo for LFLAGS, so it has been changed.
6513 13. The install script calls Exim with "-C /dev/null" in order to find the
6514 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6515 to be output. However, since Exim outputs its version number before the
6516 error, it didn't break the script. It just looked ugly. I fixed this by
6517 always allowing "-C /dev/null" if the caller is root.
6519 14. Ignore overlarge ACL variable number when reading spool file - insurance
6520 against a later release with more variables having written the file.
6522 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6523 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6524 this, as well as the form without the "IPv6" on the front.
6526 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6527 OS/Makefile-Darwin file.
6529 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6530 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6532 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6533 variable. However, some OS use ints for inodes. I've added cast to long int
6534 to get rid of the compiler warning.
6536 19. I had forgotten to lock out "/../" in configuration file names when
6537 ALT_CONFIG_PREFIX was set.
6539 20. Routers used for verification do not need to specify transports. However,
6540 if such a router generated a host list, and callout was configured, Exim
6541 crashed, because it could not find a port number from the (non-existent)
6542 transport. It now assumes port 25 in this circumstance.
6544 21. Added the -t option to exigrep.
6546 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6547 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6548 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6549 them is set, LOOKUP_LSEARCH is forced.
6551 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6553 24. Added sender and host information to the "rejected by local_scan()" log
6554 line; previously there was no indication of these.
6556 25. Added .include_if_exists.
6558 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6559 Linux. It turns out that not all file systems support this. Apparently some
6560 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6561 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6562 not supported on the file descriptor, is now ignored when Exim is trying to
6563 sync a directory. This applies only to Linux.
6565 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6567 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6568 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6571 29. Added support for maildirsize files from supplied patch (modified a bit).
6573 30. The use of :fail: followed by an empty string could lead Exim to respond to
6574 sender verification failures with (e.g.):
6576 550 Verification failed for <xxx>
6577 550 Sender verify failed
6579 where the first response line was missing the '-' that indicates it is not
6580 the final line of the response.
6582 31. The loop for finding the name of the user that called Exim had a hardwired
6583 limit of 10; it now uses the value of finduser_retries, which is used for
6584 all other user lookups.
6586 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6589 33. Exim was neglecting to zero errno before one call of strtol() when
6590 expanding a string and expecting an integer value. On some systems this
6591 resulted in spurious "integer overflow" errors. Also, it was casting the
6592 result into an int without checking.
6594 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6595 did not work. The code looks as if it has *never* worked, though it appears
6596 to have been documented since at least release 1.62. I have made it work.
6598 35. The "timeout_DNS" error in retry rules, also documented since at least
6599 1.62, also never worked. As it isn't clear exactly what this means, and
6600 clearly it isn't a major issue, I have abolished the feature by treating it
6601 as "timeout", and writing a warning to the main and panic logs.
6603 36. The display of retry rules for -brt wasn't always showing the error code
6606 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6607 timeout_connect_A, timeout_connect_MX.
6609 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6610 to the empty sender.
6612 39. The daemon was not analysing the content of -oX till after it had closed
6613 stderr and disconnected from the controlling terminal. This meant that any
6614 syntax errors were only noted on the panic log, and the return code from
6615 the command was 0. By re-arranging the code a little, I've made the
6616 decoding happen first, so such errors now appear on stderr, and the return
6617 code is 1. However, the actual setting up of the sockets still happens in
6618 the disconnected process, so errors there are still only recorded on the
6621 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6622 connections (as happens on some IP stacks) was logged at start up time as
6623 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6624 it from "IPv6 and IPv4", which means that two separate sockets are being
6627 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6628 says whether AF_INET or AF_INET6 was passed as an argument.
6630 42. Exiwhat output was messed up when time zones were included in log
6633 43. Exiwhat now gives more information about the daemon's listening ports,
6634 and whether -tls-on-connect was used.
6636 44. The "port" option of the smtp transport is now expanded.
6638 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6639 silently ignored. Now an error message is written to the main and panic
6642 46. There's a new ACL modifier called "logwrite" which writes to a log file
6643 as soon as it is encountered.
6645 47. Added $local_user_uid and $local_user_gid at routing time.
6647 48. Exim crashed when trying to verify a sender address that was being
6650 49. Exim was recognizing only a space character after ".include". It now also
6651 recognizes a tab character.
6653 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6654 extracting the relevant information from the specification. The man page no
6655 longer contains scrambled data for the -d option, and I've added a section
6656 at the front about calling Exim under different names.
6658 51. Added "extra_headers" argument to the "mail" command in filter files.
6660 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6663 53. Installed eximstats 1.29.
6665 54. Added transport_filter_timeout as a generic transport option.
6667 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6668 Cc: header lines. This was required by RFC 822, but it not required by RFC
6671 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6672 incoming messages that did not have them. Now it does so only if the
6673 message originates locally, that is, if there is no associated remote host
6674 address. When Resent- header lines are present, this applies to the Resent-
6675 lines rather than the non-Resent- lines.
6677 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6678 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6680 58. Messages for configuration errors now include the name of the main
6681 configuration file - useful now that there may be more than one file in a
6682 list (.included file names were always shown).
6684 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6685 for those rare installations that do not start the daemon as root or run it
6686 setuid root. I've cut out the call to initgroups() if the daemon is not
6689 60. The Exim user and group can now be bound into the binary as text strings
6690 that are looked up at the start of Exim's processing.
6692 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6694 62. Added $mailstore_basename variable.
6696 63. Installed patch to sieve.c from Michael Haardt.
6698 64. When Exim failed to open the panic log after failing to open the main log,
6699 the original message it was trying to log was written to stderr and debug
6700 output, but if they were not available (the usual case in production), it
6701 was lost. Now it is written to syslog before the two lines that record the
6702 failures to open the logs.
6704 65. Users' Exim filters run in subprocesses under the user's uid. It is
6705 possible for a "deliver" command or an alias in a "personal" command to
6706 provoke an address rewrite. If logging of address rewriting is configured,
6707 this fails because the process is not running as root or exim. There may be
6708 a better way of dealing with this, but for the moment (because 4.30 needs
6709 to be released), I have disabled address rewrite logging when running a
6710 filter in a non-root, non-exim process.
6716 1. The buildconfig auxiliary program wasn't quoting the value set for
6717 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6718 not defined. This bug was masked in 4.22 by the effect that was fixed in
6721 2. Some messages that were rejected after a message id was allocated were
6722 shown as "incomplete" by exigrep. It no longer does this for messages that
6723 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6725 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6726 have allow_domain_literals set, the ID did not get logged in the <= line.
6727 Domain literals are now always recognized in Message-ID: header lines.
6729 4. The first argument for a ${extract expansion item is the key name or field
6730 number. Leading and trailing spaces in this item were not being ignored,
6731 causing some misleading effects.
6733 5. When deliver_drop_privilege was set, single queue runner processes started
6734 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6735 same command in the process it spins off) were not dropping privilege.
6737 6. When the daemon running as "exim" started a queue runner, it always
6738 re-executed Exim in the spun-off process. This is a waste of effort when
6739 deliver_drop_privilege is set. The new process now just calls the
6740 queue-runner function directly.
6746 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6749 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6750 ignored. Though the use of -oP was forcing the writing of a pid file, it
6751 was always written to the default place.
6753 3. If the message "no IP address found for host xxxx" is generated during
6754 incoming verification, it is now followed by identification of the incoming
6755 connection (so you can more easily find what provoked it).
6757 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6759 5. Added some features to "harden" Exim a bit more against certain attacks:
6761 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6762 be put in Local/Makefile. This is like the never_users runtime option,
6763 but it cannot be overridden. The default setting is "root".
6765 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6766 prefix string with which any file named in a -C command line option
6769 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6770 is retained for -C and -D only if the caller of Exim is root. Without
6771 it, the exim user may also use -C and -D and retain privilege.
6773 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6774 command line option is disabled.
6776 6. Macro names set by the -D option must start with an upper case letter, just
6777 like macro names defined in the configuration file.
6779 7. Added "dereference=" facility to LDAP.
6781 8. Two instances of the typo "uknown" in the source files are fixed.
6783 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6784 the Configure-Makefile script screwed up while processing it.
6786 10. Incorporated PCRE 4.4.
6788 11. The SMTP synchronization check was not operating right at the start of an
6789 SMTP session. For example, it could not catch a HELO sent before the client
6790 waited for the greeting. There is now a check for outstanding input at the
6791 point when the greeting is written. Because of the duplex, asynchronous
6792 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6793 way, but not yet received, when the check is performed.
6795 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6796 on TCP/IP sockets, because this apparently causes some broken clients to
6799 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6800 unchanged) from the Cygwin maintainer.
6802 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6803 was supported instead of testing for mbx. Effectively a typo.
6805 15. The spa authenticator server code was not checking that the input it
6806 received was valid base64.
6808 16. The debug output line for the "set" modifier in ACLs was not showing the
6809 name of the variable that was being set.
6811 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6813 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6814 lookup on its own. Something else was needed to trigger the lookup. For
6815 example, a match in host_lookup or the need for a host name in a host list.
6816 Now, if $sender_host_name is referenced and the host name has not yet been
6817 looked up, a lookup is performed. If the lookup fails, the variable remains
6818 empty, and $host_lookup_failed is set to "1".
6820 19. Added "eqi" as a case-independent comparison operator.
6822 20. The saslauthd authentication condition could segfault if neither service
6823 nor realm was specified.
6825 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6826 error message that was logged was misleading, and incoming SMTP
6827 connections were dropped. The message is now more accurate, and temporary
6828 errors are given to SMTP connections.
6830 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6831 (except RSET) with a 421 error, until QUIT is received. However, it was
6832 failing to send a response to QUIT.
6834 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6835 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6836 ACL is now run after the helo_try_verify_hosts code.
6838 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6839 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6840 case-independent, so other case variants are also recognized). Apparently
6841 some systems use these upper case variants.
6843 25. If more than two messages were waiting for the same host, and a transport
6844 filter was specified for the transport, Exim sent two messages over the
6845 same TCP/IP connection, and then failed with "socket operation on non-
6846 socket" when it tried to send the third.
6848 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6850 27. The extern definition of crypt16() in expand.c was not being excluded when
6851 the OS had its own crypt16() function.
6853 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6854 as a preferred synonym for return_size_limit, both as an option and as an
6857 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6859 30. Changed the default configuration ACL to relax the local part checking rule
6860 for addresses that are not in any local domains. For these addresses,
6861 slashes and pipe symbols are allowed within local parts, but the sequence
6862 /../ is explicitly forbidden.
6864 31. SPA server authentication was not clearing the challenge buffer before
6867 32. log_message in a "warn" ACL statement was writing to the reject log as
6868 well as to the main log, which contradicts the documentation and doesn't
6869 seem right (because no rejection is happening). So I have stopped it.
6871 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6872 However, I am unable to do any testing of this.
6874 34. Fixed an infelicity in the appendfile transport. When checking directories
6875 for a mailbox, to see if any needed to be created, it was accidentally
6876 using path names with one or more superfluous leading slashes; tracing
6877 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6879 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6880 contents are added to the log line that is written for every discarded
6881 recipient. (Previously a log_message setting was ignored.)
6883 36. The ${quote: operator now quotes the string if it is empty.
6885 37. The install script runs exim in order to find its version number. If for
6886 some reason other than non-existence or emptiness, which it checks, it
6887 could not run './exim', it was installing it with an empty version number,
6888 i.e. as "exim-". This error state is now caught, and the installation is
6891 38. An argument was missing from the function that creates an error message
6892 when Exim fails to connect to the socket for saslauthd authentication.
6893 This could cause Exim to crash, or give a corrupted message.
6895 39. Added isip, isip4, and isip6 to ${if conditions.
6897 40. The ACL variables $acl_xx are now saved with the message, and can be
6898 accessed later in routers, transports, and filters.
6900 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6901 strings in the file are not string-expanded.
6903 42. If a MAIL command specified a SIZE value that was too large to fit into an
6904 int variable, the check against message_size_limit failed. Such values are
6905 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6906 day this will have to be increased, but I don't think I want to be around
6907 when emails are that large.
6914 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6915 iconv() is not standard in FreeBSD.
6917 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6918 IPv6 enabled. The observed symptom was a segmentation fault on return from
6919 the function os_common_find_running_interfaces() in src/os.c.
6921 3. In the check_special_case() function in daemon.c I had used "errno" as an
6922 argument name, which causes warnings on some systems. This was basically a
6923 typo, since it was named "eno" in the comments!
6925 4. The code that waits for the clock to tick (at a resolution of some fraction
6926 of a second) so as to ensure message-id uniqueness was always waiting for
6927 at least one whole tick, when it could have waited for less. [This is
6928 almost certainly not relevant at current processor speeds, where it is
6929 unlikely to ever wait at all. But we try to future-proof.]
6931 5. The function that sleeps for a time interval that includes fractions of a
6932 second contained a race. It did not block SIGALRM between setting the
6933 timer, and suspending (a couple of lines later). If the interval was short
6934 and the sigsuspend() was delayed until after it had expired, the suspension
6935 never ended. On busy systems this could lead to processes getting stuck for
6938 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6939 process, before it forks any delivery processes. The open lookup caching
6940 mechanism meant that the open file or database connection was passed into
6941 the delivery process. The problem was that delivery processes always tidy
6942 up cached lookup data. This could cause a problem for the next delivery
6943 process started by the queue runner, because the external queue runner
6944 process does not know about the closure. So the next delivery process
6945 still has data in the lookup cache. In the case of a file lookup, there was
6946 no problem because closing a file descriptor in a subprocess doesn't affect
6947 the parent. However, if the lookup was caching a connection to a database,
6948 the connection was closed, and the second delivery process was likely to
6949 see errors such as "PGSQL: query failed: server closed the connection
6950 unexpectedly". The problem has been fixed by closing all cached lookups
6951 in a queue runner before running a delivery process.
6953 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6954 seem to have the "const" qualifier which it has on other OS. I've
6957 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6958 *of the same type* (client or server) with the same public name that an
6959 error should be diagnosed.
6961 9. When Exim looked up a host name for an IP address, but failed to find the
6962 original IP address when looking up the host name (a safety check), it
6963 output the message "<ip address> does not match any IP for NULL", which was
6964 confusing, to say the least. The bug was that the host name should have
6965 appeared instead of "NULL".
6967 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6968 user that is built into the binary, and the -C or -D options is used, root
6969 privilege is dropped before the configuration file is read. In addition,
6970 logging is switched to stderr instead of the normal log files. If the
6971 configuration then re-defines the Exim user, the unprivileged environment
6972 is probably not what is expected, so Exim logs a panic warning message (but
6975 However, if deliver_drop_privilege is set, the unprivileged state may well
6976 be exactly what is intended, so the warning has been cut out in that case,
6977 and Exim is allowed to try to write to its normal log files.
6983 1. smtp_return_error_details was not giving details for temporary sender
6984 or receiver verification errors.
6986 2. Diagnose a configuration error if two authenticators have the same public
6989 3. Exim used not to create the message log file for a message until the first
6990 delivery attempt. This could be confusing when incoming messages were held
6991 for policy or load reasons. The message log file is now created at the time
6992 the message is received, and an initial "Received" line is written to it.
6994 4. The automatically generated man page for command line options had a minor
6995 bug that caused no ill effects; however, a more serious problem was that
6996 the procedure for building the man page automatically didn't always
6997 operate. Consequently, release 4.20 contains an out-of-date version. This
6998 shouldn't happen again.
7000 5. When building Exim with embedded Perl support, the script that builds the
7001 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7002 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7004 6. The freeze_tell option was not being used for messages that were frozen on
7005 arrival, either by an ACL or by local_scan().
7007 7. Added the smtp_incomplete_transaction log selector.
7009 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7010 was accepting AUTH without a new EHLO.
7012 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7013 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7016 10. Logging of TCP/IP connections (when configured) now happens in the main
7017 daemon process instead of the child process, so that the TCP/IP connection
7018 count is more accurate (but it can never be perfect).
7020 11. The use of "drop" in a nested ACL was not being handled correctly in the
7021 outer ACL. Now, if condition failure induced by the nested "drop" causes
7022 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7023 or "require"), the connection is dropped.
7025 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7026 that yield "discard" can now be used with an "accept" or a "discard" verb,
7027 but an error is generated for any others (because I can't see a useful way
7028 to define what should happen).
7030 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7031 are now processed in the same way as lines in the Exim configuration file.
7032 In particular, continuation lines are supported.
7034 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7036 15. Added -ti meaning -t -i.
7038 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7039 domains, and warn by logging if others are found.
7041 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7042 returned from GIFCONF when Exim is trying to find the list of interfaces on
7043 a host. The code in os.c has been modified to copy each ifreq to an aligned
7044 structure in all cases.
7046 Also, in some cases, the returned ifreq's were being copied to a 'struct
7047 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7048 means the last couple of bytes of an IPv6 address could be chopped if the
7049 ifreq contained only a normal sockaddr (14 bytes storage).
7051 18. Named domain lists were not supported in the hosts_treat_as_local option.
7052 An entry such as +xxxx was not recognized, and was treated as a literal
7055 19. Ensure that header lines added by a DATA ACL are included in the reject log
7056 if the ACL subsequently rejects the message.
7058 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7059 MD5 (which is deprecated).
7061 21. When testing a filter file using -bf, Exim was writing a message when it
7062 took the sender from a "From " line in the message, but it was not doing so
7063 when it took $return_path from a Return-Path: header line. It now does.
7065 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7066 with a valid header line field name (a series of printing characters
7067 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7069 23. Changed "disc" in the source to "disk" to conform to the documentation and
7070 the book and for uniformity.
7072 24. Ignore Sendmail's -Ooption=value command line item.
7074 25. When execve() failed while trying to run a command in a pipe transport,
7075 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7076 could be confused with a return value of 69 from the command itself. This
7077 has been changed to 127, the value the shell returns if it is asked to run
7078 a non-existent command. The wording for the related log line suggests a
7079 non-existent command as the problem.
7081 26. If received_header_text expands to an empty string, do not add a Received:
7082 header line to the message. (Well, it adds a token one on the spool, but
7083 marks it "old" so that it doesn't get used or transmitted.)
7085 27. Installed eximstats 1.28 (addition of -nt option).
7087 28. There was no check for failure on the call to getsockname() in the daemon
7088 code. This can fail if there is a shortage of resources on the system, with
7089 ENOMEM, for example. A temporary error is now given on failure.
7091 29. Contrary to the C standard, it seems that in some environments, the
7092 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7093 program. Exim now does this explicitly; it affects the formatting of
7094 timestamps using strftime().
7096 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7097 complaints. I've now initialized all the variables, to avoid this.
7099 32. Header lines added by a system filter were not being "seen" during
7100 transport-time rewrites.
7102 33. The info_callback() function passed to OpenSSL is set up with type void
7103 (*)(SSL *, int, int), as described somewhere. However, when calling the
7104 function (actually a macro) that sets it up, the type void(*)() is
7105 expected. I've put in a cast to prevent warnings from picky compilers.
7107 34. If a DNS black list lookup found a CNAME record, but there were no A
7108 records associated with the domain it pointed at, Exim crashed.
7110 35. If a DNS black list lookup returned more than one A record, Exim ignored
7111 all but the first. It now scans all returned addresses if a particular IP
7112 value is being sought. In this situation, the contents of the
7113 $dnslist_value variable are a list of all the addresses, separated by a
7116 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7117 to accept a host name and all its aliases if the forward lookup for any of
7118 them yielded the IP address of the incoming connection. Now it accepts only
7119 those names whose forward lookup yields the correct IP address. Any other
7120 names are discarded. This closes a loophole whereby a rogue DNS
7121 administrator could create reverse DNS records to break through a
7122 wildcarded host restriction in an ACL.
7124 37. If a user filter or a system filter that ran in a subprocess used any of
7125 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7126 the wrong values were passed to the pipe command ($thisaddress had the
7127 value of $0, $0 had the value of $1, etc). This bug was introduced by
7128 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7130 38. Improved the line breaking for long SMTP error messages from ACLs.
7131 Previously, if there was no break point between 40 and 75 characters, Exim
7132 left the rest of the message alone. Two changes have been made: (a) I've
7133 reduced the minimum length to 35 characters; (b) if it can't find a break
7134 point between 35 and 75 characters, it looks ahead and uses the first one
7135 that it finds. This may give the occasional overlong line, but at least the
7136 remaining text gets split now.
7138 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7139 file descriptors might be low, and that setting 1000 would always raise it.
7140 It turns out that in some environments, the limit is already over 1000 and
7141 that lowering it causes trouble. So now Exim takes care not to decrease it.
7143 40. When delivering a message, the value of $return_path is set to $sender_
7144 address at the start of routing (routers may change the value). By an
7145 oversight, this default was not being set up when an address was tested by
7146 -bt or -bv, which affected the outcome if any router or filter referred to
7149 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7150 log only when "message" or "log_message" are set. However, if one of the
7151 conditions was an address verification, or a call to a nested ACL, the
7152 messages generated by the underlying test were being passed through. This
7153 no longer happens. The underlying message is available in $acl_verify_
7154 message for both "message" and "log_message" expansions, so it can be
7155 passed through if needed.
7157 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7158 new expansion $bh_ to give the encoded byte string without charset
7159 translation. Translation happens only if iconv() is available; HAVE_ICONV
7160 indicates this at build time. HEADERS_CHARSET gives the charset to
7161 translate to; headers_charset can change it in the configuration, and
7162 "headers charset" can change it in an individual filter file.
7164 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7165 that creates RFC 2047 encoded "words" labels them as that charset instead
7166 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7167 expansion operator; (ii) when Exim creates a From: line for a local
7168 message; (iii) when a header line is rewritten to include a "phrase" part.
7170 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7171 buggy, causing it to skip the first lines of messages whose message ID
7172 ended in 'D'. This would not have bitten before Exim release 4.14, because
7173 message IDs were unlikely to end in 'D' before then. The effect was to have
7174 incorrect size information for certain domains.
7176 45. #include "config.h" was missing at the start of the crypt16.c module. This
7177 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7180 46. If there was a timeout during a "random" callout check, Exim treated it as
7181 a failure of the random address, and carried on sending RSET and the real
7182 address. If the delay was just some slowness somewhere, the response to the
7183 original RCPT would be taken as a response to RSET and so on, causing
7184 mayhem of various kinds.
7186 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7187 when I implemented it. It didn't allow for the fact that some option values
7188 may legitimately be negative (e.g. size_addition), and it didn't even do
7189 the right test for positive values.
7191 48. Domain names in DNS records are case-independent. Exim always looks them up
7192 in lower case. Some resolvers return domain names in exactly the case they
7193 appear in the zone file, that is, they may contain uppercase letters. Not
7194 all resolvers do this - some return always lower case. Exim was treating a
7195 change of case by a resolver as a change of domain, similar to a widening
7196 of a domain abbreviation. This triggered its re-routing code and so it was
7197 trying to route what was effectively the same domain again. This normally
7198 caused routing to fail (because the router wouldn't handle the domain
7199 twice). Now Exim checks for this case specially, and just changes the
7200 casing of the domain that it ultimately uses when it transmits the message
7203 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7206 50. If a filter generated a file delivery with a non-absolute name (possible if
7207 no home directory exists for the router), the forbid_file option was not
7210 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7211 the existing equality matching.
7213 52. Exim was using ints instead of ino_t variables in some places where it was
7214 dealing with inode numbers.
7216 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7217 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7218 called TMPDIR, and if it finds it is different, it changes its value.
7220 54. The smtp_printf() function is now made available to local_scan() so
7221 additional output lines can be written before returning. There is also an
7222 smtp_fflush() function to enable the detection of a dropped connection.
7223 The variables smtp_input and smtp_batched_input are exported to
7226 55. Changed the default runtime configuration: the message "Unknown user"
7227 has been removed from the ACL, and instead placed on the localuser router,
7228 using the cannot_route_message feature. This means that any verification
7229 failures that generate their own messages won't get overridden. Similarly,
7230 the "Unrouteable address" message that was in the ACL for unverifiable
7231 relay addresses has also been removed.
7233 56. Added hosts_avoid_esmtp to the smtp transport.
7235 57. The exicyclog script was not checking for the esoteric option
7236 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7237 will work only if exicyclog is run under the appropriate euid.
7239 58. Following a discussion on the list, the rules by which Exim recognises line
7240 endings on incoming messages have been changed. The -dropcr and drop_cr
7241 options are now no-ops, retained only for backwards compatibility. The
7242 following line terminators are recognized: LF CRLF CR. However, special
7243 processing applies to CR:
7245 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7246 nor a local message in the state where . is a terminator.
7248 (ii) If a bare CR is encountered in a header line, an extra space is added
7249 after the line terminator so as not to end the header. The reasoning
7250 behind this is that bare CRs in header lines are most likely either
7251 to be mistakes, or people trying to play silly games.
7253 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7254 was being incorrectly given as 18 bytes larger than it should have been.
7255 This is a VOB (very old bug).
7257 60. This may never have affected anything current, but just in case it has:
7258 When the local host is found other than at the start of a list of hosts,
7259 the local host, those with the same MX, and any that follow, are discarded.
7260 When the list in question was part of a longer list of hosts, the following
7261 hosts (not currently being processed) were also being discarded. This no
7262 longer happens. I'm not sure if this situation could ever has previously
7265 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7268 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7269 header. This is something that is recommended in a new Internet Draft, and
7270 is something that is documented as being done by Sendmail. There are two
7271 possible values. For messages generated by the autoreply transport, Exim
7274 Auto-Submitted: auto-replied
7276 whereas for all other generated messages (e.g. bounces) it adds
7278 Auto-Submitted: auto-generated
7280 63. The "personal" condition in filters now includes a test for the
7281 Auto-Submitted: header. If it contains the string "auto-" the message it
7282 not considered personal.
7284 64. Added rcpt_include_affixes as a generic transport option.
7286 65. Added queue_only_override (default true).
7288 66. Added the syslog_duplication option.
7290 67. If what should have been the first header line of a message consisted of
7291 a space followed by a colon, Exim was mis-interpreting it as a header line.
7292 It isn't of course - it is syntactically invalid and should therefore be
7293 treated as the start of the message body. The misbehaviour could have
7294 caused a number of strange effects, including loss of data in subsequent
7295 header lines, and spool format errors.
7297 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7298 client host had authenticated. This control can now be exercised by an ACL
7299 for more flexibility.
7301 69. By default, callouts do not happen when testing with -bh. There is now a
7302 variant, -bhc, which does actually run the callout code, including
7303 consulting and updating the callout cache.
7305 70. Added support for saslauthd authentication, courtesy of Alexander
7308 71. If statvfs() failed on the spool or log directories while checking their
7309 size for availability, Exim confusingly gave the error "space shortage".
7310 Furthermore, in debugging mode it crashed with a floating point exception.
7311 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7312 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7313 serious problem, Exim now writes to the main and panic logs when this
7314 happens, with details of the failure. It then refuses to accept the
7315 incoming message, giving the message "spool directory problem" or "log
7316 directory problem" with a 421 code for SMTP messages.
7318 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7319 0, 1, and 2 exist, because some OS complain for execs without them (see
7320 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7321 descriptors. However, the code omitted to check that the open succeeded,
7322 causing mysterious errors if for some reason the permissions on /dev/null
7323 got screwed. Now Exim writes a message to the main and panic logs, and
7324 bombs out if it can't open /dev/null.
7326 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7327 interact so that it is all more flexible. It is supposed to remain
7328 backwards compatible. Also added extra_local_interfaces.
7330 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7331 to bomb out with an assertion failure - to the client this appears as a
7332 connection drop. This problem occurs in the part of the code that was taken
7333 from the Samba project. Fortunately, the assertion is in a very simple
7334 function, so I have fixed this by reproducing the function inline in the
7335 one place where it is called, and arranging for authentication to fail
7336 instead of killing the process with assert().
7338 75. The SPA client code was not working when the server requested OEM rather
7339 than Unicode encoding.
7341 76. Added code to make require_files with a specific uid setting more usable in
7342 the case where statting the file as root fails - usually a non-root-mounted
7343 NFS file system. When this happens and the failure is EACCES, Exim now
7344 forks a subprocess and does the per-uid checking as the relevant uid.
7346 77. Added process_log_path.
7348 78. If log_file_path was not explicitly set, a setting of check_log_space or
7349 check_log_inodes was ignored.
7351 79. If a space check for the spool or log partitions fails, the incident is now
7352 logged. Of course, in the latter case the data may get lost...
7354 80. Added the %p formatting code to string_format() so that it can be used to
7355 print addresses in debug_print(). Adjusted all the address printing in the
7356 debugging in store.c to use %p rather than %d.
7358 81. There was a concern that a line of code in smtp_in.c could overflow a
7359 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7360 initially expressed, the concern was not well-founded, because trailing
7361 spaces are removed early. However, if the trailing spaces were followed by
7362 a NULL, they did not get removed, so the overflow was possible. Two fixes
7365 (a) I re-wrote the offending code in a cleaner fashion.
7366 (b) If an incoming SMTP command contains a NULL character, it is rejected
7369 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7370 runs initgroups(), so that if the Exim user is in any additional groups,
7371 they will be used during message reception.
7377 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob