1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
12 buffer overrun for (non-chunking) other transports.
14 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
15 TLS1.3, means that a server rejecting a client certificate is not visible
16 to the client until the first read of encrypted data (typically the
17 response to EHLO). Add detection for that case and treat it as a failed
18 TLS connection attempt, so that the normal retry-in-clear can work (if
21 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
22 and/or domain. Found and fixed by Jason Betts.
24 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
25 configuration). If a CNAME target was not a wellformed name pattern, a
28 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
29 the OS reports them interleaved with other addresses.
31 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
32 used both for input and for a verify callout, both encrypted, SMTP
33 responses being sent by the server could be lost. This resulted in
34 dropped connections and sometimes bounces generated by a peer sending
37 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
38 string. Previously it was possible to cause undefined behaviour in a
39 library routine (usually a crash). Found by "zerons".
41 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
49 JH/01 Remove code calling the customisable local_scan function, unless a new
50 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
52 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
53 non-signal-safe functions being used.
55 JH/03 Bug 2269: When presented with a received message having a stupidly large
56 number of DKIM-Signature headers, disable DKIM verification to avoid
57 a resource-consumption attack. The limit is set at twenty.
59 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
60 report of oldest_pass in ${authres } in consequence, and separate out
61 some descriptions of reasons for verification fail.
63 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
64 files in the spool were present and unlocked. A queue-runner could spot
65 them, resulting in a duplicate delivery. Fix that by doing the unlock
66 after the unlink. Investigation by Tim Stewart. Take the opportunity to
67 add more error-checking on spoolfile handling while that code is being
70 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
71 No known attacks, no CVE, this is defensive hardening.
73 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
74 a queue-runner could start a delivery while other operations were ongoing.
75 Cutthrough delivery was a common victim, resulting in duplicate delivery.
76 Found and investigated by Tim Stewart. Fix by using the open message data
77 file handle rather than opening another, and not locally closing it (which
78 releases a lock) for that case, while creating the temporary .eml format
79 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
81 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
82 $sender_verify_failure/$recipient_verify_failure to "random".
84 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
87 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
88 Previously this would segfault.
90 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
93 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
94 like zero, since the resolver should be doing this for us, But we need one
95 as a CNAME but no MX presence gets the CNAME returned; we need to check
96 that doesn't point to an MX to declare it "no MX returned" rather than
97 "error, loop". A new main option is added so the older capability of
98 following some limited number of chain links is maintained.
100 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
102 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
103 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
104 now-deprecated earlier definitions used only specified the range up to TLS
105 1.2 (in the older-version library docs).
107 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
109 JH/15 Rework TLS client-side context management. Stop using a global, and
110 explicitly pass a context around. This enables future use of TLS for
111 connections to service-daemons (eg. malware scanning) while a client smtp
112 connection is using TLS; with cutthrough connections this is quite likely.
114 JH/16 Fix ARC verification to do AS checks in reverse order.
116 JH/17 Support a "tls" option on the ${readsocket } expansion item.
118 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
119 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
120 Previously the "utf8" would be re-prepended for every additional message.
122 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
123 Previously thery were accepted, resulting in issues when attempting to
124 forward messages to a non-supporting MTA.
126 PP/02 Let -n work with printing macros too, not just options.
128 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
129 one parent address was copied, and bogus data was used at delivery-logging
130 time. Either a crash (after delivery) or bogus log data could result.
131 Discovery and analysis by Tim Stewart.
133 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
134 Previously if the string ended mid-character, we did not insert the
135 promised '?' replacement.
137 PP/04 Documentation: current string operators work on bytes, not codepoints.
139 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
140 should pack well giving a smaller memory footprint so better caching and
141 therefore performance. Group the declarations where this can't be done so
142 that the byte-sized flag variables are not interspersed among pointer
143 variables, giving a better chance of good packing by the compiler.
145 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
146 non-null, to avoid issues with sites running BATV. Previously reports were
147 sent with an empty envelope sender so looked like bounces.
149 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
150 The ignore_error flag wasn't being returned from the filter subprocess so
151 was not set for later routers. Investigation and fix by Matthias Kurz.
153 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
154 and a msg:complete for the whole, when a message is manually removed using
155 -Mrm. Developement by Matthias Kurz, hacked on by JH.
157 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
158 a "Gnu special" function, asprintf() in the DB utility binary builds; I
159 hope that is portable enough.
161 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
162 requiring a known-CA anchor certificate; make it now rely entirely on the
163 TLSA as an anchor. Checking the name on the leaf cert against the name
164 on the A-record for the host is still done for TA (but not for EE mode).
166 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
167 included in delivery lines for non-proxied connections, when compiled with
168 SUPPORT_SOCKS and running with proxy logging enabled.
170 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
171 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
172 move the existing event to fire before the normal logging of message
173 failure so that custom logging is bracketed by normal logging.
175 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
176 msg:fail:internal event. Developement by Matthias Kurz.
178 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
179 far too small for todays use of crypto signatures stored there. Go all
180 the way to the max DNS message size of 64kB, even though this might be
181 overmuch for IOT constrained device use.
183 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
184 copy a string over itself. The library routine is documented as not
185 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
187 JH/32 For main options check_spool_space and check_inode_space, where the
188 platform supports 64b integers, support more than the previous 2^31 kB
189 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
190 the previous G, M, k.
192 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
193 $authenticated_fail_id variable on authentication failure. Previously
196 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
197 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
198 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
199 GNUTLS_SEC_PARAM_MEDIUM.
201 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
202 side. Previously we would continue as if no SNI had been received.
204 JH/36 Harden the handling of string-lists. When a list consisted of a sole
205 "<" character, which should be a list-separator specification, we walked
206 off past the nul-terimation.
208 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
209 causes) even when the retry time is not yet met. Previously they were
210 not, meaning that when (say) an account was over-quota and temp-rejecting,
211 and multiple senders' messages were queued, only one sender would get
212 notified on each configured delay_warning cycle.
214 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
216 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
217 cipher-suites, an error can be left on the stack even for a succeeding
218 accept; this results in impossible error messages when a later operation
221 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
222 return error codes indicating retry. Under TLS1.3 this becomes required.
224 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
225 it only wrote the new authenticators, resulting in a lack of tracking of
226 peer changes of ESMTP extensions until the next cache flush.
228 JH/41 Fix the loop reading a message header line to check for integer overflow,
229 and more-often against header_maxsize. Previously a crafted message could
230 induce a crash of the recive process; now the message is cleanly rejected.
232 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
233 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
239 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
240 When redis_servers is set to a list of > 1 element, and the Redis servers
241 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
242 case of MOVED into a DEFER case instead, thus moving the query onto the
243 next server in the list. For a cluster of N elements, all N servers must
244 be defined in redis_servers.
246 GF/02 Catch and remove uninitialized value warning in exiqsumm
247 Check for existence of @ARGV before looking at $ARGV[0]
249 JH/01 Replace the store_release() internal interface with store_newblock(),
250 which internalises the check required to safely use the old one, plus
251 the allocate and data copy operations duplicated in both (!) of the
252 extant use locations.
254 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
255 modifier. This matches the restriction on the commandline.
257 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
258 Previously only the last row was returned.
260 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
261 we assumed that tags in the header were well-formed, and parsed the
262 element content after inspecting only the first char of the tag.
263 Assumptions at that stage could crash the receive process on malformed
266 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
267 While running the DKIM ACL we operate on the Permanent memory pool so that
268 variables created with "set" persist to the DATA ACL. Also (at any time)
269 DNS lookups that fail create cache records using the Permanent pool. But
270 expansions release any allocations made on the current pool - so a dnsdb
271 lookup expansion done in the DKIM ACL releases the memory used for the
272 DNS negative-cache, and bad things result. Solution is to switch to the
273 Main pool for expansions.
274 While we're in that code, add checks on the DNS cache during store_reset,
275 active in the testsuite.
276 Problem spotted, and debugging aided, by Wolfgang Breyha.
278 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
279 When none of the hosts presented to a transport match an already-open
280 connection, close it and proceed with the list. Previously we would
281 queue the message. Spotted by Lena with Yahoo, probably involving
284 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
285 Previously a spurious "250 OK id=" response was appended to the proper
288 JH/08 The "support for" informational output now, which built with Content
289 Scanning support, has a line for the malware scanner interfaces compiled
290 in. Interface can be individually included or not at build time.
292 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
293 by the template makefile "src/EDITME". The "STREAM" support for an older
294 ClamAV interface method is removed.
296 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
297 rows affected is given instead).
299 JH/11 The runtime Berkeley DB library version is now additionally output by
300 "exim -d -bV". Previously only the compile-time version was shown.
302 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
303 SMTP connection. Previously, when one had more receipients than the
304 first, an abortive onward connection was made. Move to full support for
305 multiple onward connections in sequence, handling cutthrough connection
306 for all multi-message initiating connections.
308 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
309 routers. Previously, a multi-recipient message would fail to match the
310 onward-connection opened for the first recipient, and cause its closure.
312 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
313 a timeout on read on a GnuTLS initiating connection, resulting in the
314 initiating connection being dropped. This mattered most when the callout
315 was marked defer_ok. Fix to keep the two timeout-detection methods
318 JH/15 Relax results from ACL control request to enable cutthrough, in
319 unsupported situations, from error to silently (except under debug)
320 ignoring. This covers use with PRDR, frozen messages, queue-only and
323 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
325 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
326 metadata, resulting in a crash in free().
328 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
329 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
330 Broken also in d185889f4, with init system revamp.
332 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
333 Previously we abruptly closed the connection after reading a malware-
334 found indication; now we go on to read the "scan ok" response line,
337 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
338 ACL. Previously, a crash would result.
340 JH/19 Speed up macro lookups during configuration file read, by skipping non-
341 macro text after a replacement (previously it was only once per line) and
342 by skipping builtin macros when searching for an uppercase lead character.
344 JH/20 DANE support moved from Experimental to mainline. The Makefile control
345 for the build is renamed.
347 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
348 was allocated for every new TLS startup, meaning one per message. Fix
349 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
351 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
352 reported the original. Fix to report (as far as possible) the ACL
353 result replacing the original.
355 JH/23 Fix memory leak during multi-message connections using STARTTLS under
356 OpenSSL. Certificate information is loaded for every new TLS startup,
357 and the resources needed to be freed.
359 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
361 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
364 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
365 DATA response info to the (existing) per-recipient response info for
366 the "C=" log element. It can have useful tracking info from the
367 destination system. Patch from Simon Arlott.
369 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
370 length value. Previously this would segfault.
372 HS/02 Support Avast multiline protoocol, this allows passing flags to
373 newer versions of the scanner.
375 JH/28 Ensure that variables possibly set during message acceptance are marked
376 dead before release of memory in the daemon loop. This stops complaints
377 about them when the debug_store option is enabled. Discovered specifically
378 for sender_rate_period, but applies to a whole set of variables.
379 Do the same for the queue-runner and queue-list loops, for variables set
380 from spool message files. Do the same for the SMTP per-message loop, for
381 certain variables indirectly set in ACL operations.
383 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
384 as a multi-recipient message from a mailinglist manager). The coding had
385 an arbitrary cutoff number of characters while checking for more input;
386 enforced by writing a NUL into the buffer. This corrupted long / fast
387 input. The problem was exposed more widely when more pipelineing of SMTP
388 responses was introduced, and one Exim system was feeding another.
389 The symptom is log complaints of SMTP syntax error (NUL chars) on the
390 receiving system, and refused recipients seen by the sending system
391 (propating to people being dropped from mailing lists).
392 Discovered and pinpointed by David Carter.
394 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
395 replaced by the ${authresults } expansion.
397 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
399 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
400 allows proper process termination in container environments.
402 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
403 Previously the "final dot" had a newline after it; ensure it is CR,LF.
405 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
406 and "err_perm", deprecated since 4.83 when the RFC-defined words
407 "temperror" and "permerror" were introduced.
409 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
410 transport-filters or DKIM-signing. The restriction was lost in the
411 consolidation of verify-callout and delivery SMTP handling.
412 Extend the restriction to also cover ARC-signing.
414 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
415 in defer=pass mode supply a 450 to the initiator. Previously the message
418 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
419 tls_require_ciphers is used as before.
421 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
422 "pass_unscanned". Only tmpfails from the scanner are written to
423 the paniclog, as they may require admin intervention (permission
424 denied, license issues). Other scanner errors (like decompression
425 bombs) do not cause a paniclog entry.
427 JH/36 Fix reinitialisation of DKIM logging variable between messages.
428 Previously it was possible to log spurious information in receive log
431 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
432 triggered odd behaviour from Outlook Express clients.
434 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
437 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
438 since the IETF WG has not yet settled on that versus the original
439 "bare" representation.
441 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
442 Previously the millisecond value corrupted the output.
443 Fix also for syslog_pid=no and log_selector +pid, for which the pid
444 corrupted the output.
450 JH/01 Rework error string handling in TLS interface so that the caller in
451 more cases is responsible for logging. This permits library-sourced
452 string to be attached to addresses during delivery, and collapses
453 pairs of long lines into single ones.
455 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
456 during configuration. Wildcards are allowed and expanded.
458 JH/02 Rework error string handling in DKIM to pass more info back to callers.
459 This permits better logging.
461 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
462 do not close it down and have the child transport start it up again on
463 the passed-on TCP connection. Instead, proxy the child (and any
464 subsequent ones) for TLS via a unix-domain socket channel. Logging is
465 affected: the continued delivery log lines do not have any DNSSEC, TLS
466 Certificate or OCSP information. TLS cipher information is still logged.
468 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
469 identical IP addresses on different listening ports. Will also affect
472 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
473 add noisy ifdef guards to special-case this sillyness.
474 Patch from Bernd Kuhls.
476 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
477 than 255 are no longer allowed.
479 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
480 Disable the session-cache too, which might reduce our load. Since we
481 currrectly use a new context for every connection, both as server and
482 client, there is no benefit for these.
483 GnuTLS appears to not support tickets server-side by default (we don't
484 call gnutls_session_ticket_enable_server()) but client side is enabled
485 by default on recent versions (3.1.3 +) unless the PFS priority string
488 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
489 <https://reproducible-builds.org/specs/source-date-epoch/>.
491 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
492 the check for any unsuccessful recipients did not notice the limit, and
493 erroneously found still-pending ones.
495 JH/08 Pipeline CHUNKING command and data together, on kernels that support
496 MSG_MORE. Only in-clear (not on TLS connections).
498 JH/09 Avoid using a temporary file during transport using dkim. Unless a
499 transport-filter is involved we can buffer the headers in memory for
500 creating the signature, and read the spool data file once for the
501 signature and again for transmission.
503 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
504 4.77 as the kernel support then wasn't solid, having issues in 64bit
505 mode. Now, it's been long enough. Add support for FreeBSD also.
507 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
508 case where the routing stage had gathered several addresses to send to
509 a host before calling the transport for the first, we previously failed
510 to close down TLS in the old transport process before passing the TCP
511 connection to the new process. The new one sent a STARTTLS command
512 which naturally failed, giving a failed delivery and bloating the retry
513 database. Investigation and fix prototype from Wolfgang Breyha.
515 JH/12 Fix check on SMTP command input synchronisation. Previously there were
516 false-negatives in the check that the sender had not preempted a response
517 or prompt from Exim (running as a server), due to that code's lack of
518 awareness of the SMTP input buffering.
520 PP/04 Add commandline_checks_require_admin option.
521 Exim drops privileges sanely, various checks such as -be aren't a
522 security problem, as long as you trust local users with access to their
523 own account. When invoked by services which pass untrusted data to
524 Exim, this might be an issue. Set this option in main configuration
525 AND make fixes to the calling application, such as using `--` to stop
528 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
529 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
530 responses to those, into a single TLS record each way (this usually means
531 a single packet). As a side issue, smtp_enforce_sync now works on TLS
534 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
535 affects you only if you're dancing at the edge of the param size limits.
536 If you are, and this message makes sense to you, then: raise the
537 configured limit or use OpenSSL 1.1. Nothing we can do for older
540 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
541 cmdline element to get the documented default one. Previously it was
544 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
545 are unsuable for non-message-specific reasons - eg. connection timeout,
546 banner-time rejection.
548 JH/16 Fix logging of delivery remote port, when specified by router, under
551 PP/06 Repair manualroute's ability to take options in any order, even if one
552 is the name of a transport.
555 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
557 JH/17 Change the list-building routines interface to use the expanding-string
558 triplet model, for better allocation and copying behaviour.
560 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
561 Previously it was constructed the first time a possibly-matching string
562 was met in the configuration file input during startup; now it is done
565 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
566 compatible one, to avoid the (poorly documented) possibility of a config
567 file in the working directory redirecting the DB files, possibly correpting
568 some existing file. CVE-2017-10140 assigned for BDB.
570 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
571 cache-hot. Previously, although the result was properly cached, the
572 initial verify call returned a defer.
574 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
575 the main verify for receipient in uncached-mode.
577 JH/22 Retire historical build files to an "unsupported" subdir. These are
578 defined as "ones for which we have no current evidence of testing".
580 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
581 if present. Previously it was ignored.
583 JH/24 Start using specified-initialisers in C structure init coding. This is
584 a C99 feature (it's 2017, so now considered safe).
586 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
587 if was a fixed-sized field and bitmask ops via macros; it is now more
590 PP/07 GitHub PR 56: Apply MariaDB build fix.
591 Patch provided by Jaroslav Škarvada.
593 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
594 during Coverity cleanups [4.87 JH/47]
595 Diagnosis and fix provided by Michael Fischer v. Mollard.
597 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
598 the right size to place the terminating semicolon on its own folded
599 line, the header hash was calculated to an incorrect value thanks to
600 the (relaxed) space the fold became.
602 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
603 and confused the parent.
605 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
606 which could crash as a result. This could lead to undeliverable messages.
608 JH/28 Logging: "next input sent too soon" now shows where input was truncated
611 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
612 matters on fast-turnover and PID-randomising systems, which were getting
613 out-of-order delivery.
615 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
616 a possibly-overlapping copy. The symptom was that "Remote host closed
617 connection in response to HELO" was logged instead of the actual 4xx
620 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
621 Previously only that bufferd was discarded, resulting in SYMTP command
624 JH/32 DKIM: when a message has multiple signatures matching an identity given
625 in dkim_verify_signers, run the dkim acl once for each. Previously only
626 one run was done. Bug 2189.
628 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
629 "panic the current process" to "deliberately defer". The panic log is
630 still written with the problem list name; the mail and reject logs now
631 get a temp-reject line for the message that was being handled, saying
632 something like "domains check lookup or other defer". The SMTP 451
633 message is still "Temporary local problem".
635 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
636 A crafted sequence of BDAT commands could result in in-use memory beeing
637 freed. CVE-2017-16943.
639 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
640 from SMTP input. Previously it was always done; now only done for DATA
641 and not BDAT commands. CVE-2017-16944.
643 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
644 to the message (such as an overlong header line). Previously this was
645 not done and we did not exit BDAT mode. Followon from the previous item
646 though a different problem.
652 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
653 than -2003 did; needs libidn2 in addition to libidn.
655 JH/02 The path option on a pipe transport is now expanded before use.
657 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
658 Patch provided by "Björn", documentation fix added too.
660 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
661 missing a wire-to-host endian conversion.
663 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
664 close after a BDAT command line could be taken as a following command,
665 giving a synch failure. Fix by only checking for synch immediately
666 before acknowledging the chunk.
668 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
669 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
670 macro. Patches provided by Josh Soref.
672 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
673 Previously we did not; the RFC seems ambiguous and VRFY is not listed
674 by IANA as a service extension. However, John Klensin suggests that we
677 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
678 the dkim code may be unix-mode line endings rather than smtp wire-format
679 CRLF, so prepend a CR to any bare LF.
681 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
682 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
684 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
685 the first were themselves being wrongly included in the feed into dkim
686 processing; with most chunk sizes in use this resulted in an incorrect
687 body hash calculated value.
689 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
690 DKIM signature block, for verification. Although advised against by
691 standards it is specifically not ruled illegal.
693 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
695 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
696 missing a body hash (the bh= tag).
698 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
699 It seems that HAProxy sends the Proxy Protocol information in clear and
700 only then does a TLS startup, so do the same.
702 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
703 TCP connections (such as for Spamd) unless the daemon successfully set
704 Fast Open mode on its listening sockets. This fixes breakage seen on
705 too-old kernels or those not configured for Fast Open, at the cost of
706 requiring both directions being enabled for TFO, and TFO never being used
707 by non-daemon-related Exim processes.
709 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
710 endings, at least on the first header line. Try to canonify any that get
711 past that check, despite the cost.
713 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
714 now limited to an arbitrary five deep, while parsing addresses with the
715 strip_excess_angle_brackets option enabled.
717 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
718 instead leave the unprompted TLS handshake in socket buffer for the
719 TLS library to consume.
721 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
723 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
725 JH/16 Drop variables when they go out of scope. Memory management drops a whole
726 region in one operation, for speed, and this leaves assigned pointers
727 dangling. Add checks run only under the testsuite which checks all
728 variables at a store-reset and panics on a dangling pointer; add code
729 explicitly nulling out all the variables discovered. Fixes one known
730 bug: a transport crash, where a dangling pointer for $sending_ip_address
731 originally assigned in a verify callout, is re-used.
733 PP/06 Drop '.' from @INC in various Perl scripts.
735 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
737 PP/08 Reduce a number of compilation warnings under clang; building with
738 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
739 should be warning-free.
741 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
743 HS/01 Fix portability problems introduced by PP/08 for platforms where
744 realloc(NULL) is not equivalent to malloc() [SunOS et al].
746 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
747 chunk. This allows us to accept broken chunked messages. We need a more
748 general solution here.
750 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
751 already-broken messages in the queue.
753 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
755 JH/19 Fix reference counting bug in routing-generated-address tracking.
761 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
762 supports it and a size is available (ie. the sending peer gave us one).
764 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
765 years of being deprecated). The replacements are the ACLs
766 acl_smtp_mime and acl_not_smtp_mime.
768 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
769 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
770 or even an in-clear connection were permitted. Now, if the host lookup
771 was dnssec and dane was requested then the host is only used if the
772 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
773 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
774 if one fails this test.
775 This means that a poorly-configured remote DNS will make it incommunicado;
776 but it protects against a DNS-interception attack on it.
778 JH/04 Bug 1810: make continued-use of an open smtp transport connection
779 non-noisy when a race steals the message being considered.
781 JH/05 If main configuration option tls_certificate is unset, generate a
782 self-signed certificate for inbound TLS connections.
784 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
785 in rewrites and routers.
787 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
788 and logged a warning sing 4.83; now they are a configuration file error.
790 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
791 (lacking @domain). Apply the same qualification processing as RCPT.
793 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
795 JH/10 Support ${sha256:} applied to a string (as well as the previous
798 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
799 a cutthrough deliver is pending, as we always want to make a connection.
800 This also avoids re-routing the message when later placing the cutthrough
801 connection after a verify cache hit.
802 Do not update it with the verify result either.
804 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
805 when routing results in more than one destination address.
807 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
808 signing (which inhibits the cutthrough capability). Previously only
809 the presence of an option was tested; now an expansion evaluating as
810 empty is permissible (obviously it should depend only on data available
811 when the cutthrough connection is made).
813 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
814 the relevant preceding SMTP command did not note the pipelining mode.
816 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
817 Previously they were not counted.
819 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
820 as one having no matching records. Previously we deferred the message
821 that needed the lookup.
823 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
824 distinguished as "(=".
826 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
827 for missing MX records. Previously it only worked for missing A records.
829 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
831 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
832 after the data-go-ahead and data-ack. Patch from Jason Betts.
834 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
835 even for a "none" policy. Patch from Tony Meyer.
837 JH/22 Fix continued use of a connection for further deliveries. If a port was
838 specified by a router, it must also match for the delivery to be
841 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
842 When one of the recipients of a message was unsuitable for the connection
843 (has no matching addresses), we lost track of needing to mark it
844 deferred. As a result mail would be lost.
846 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
848 JH/25 Decoding ACL controls is now done using a binary search; the source code
849 takes up less space and should be simpler to maintain. Merge the ACL
850 condition decode tables also, with similar effect.
852 JH/26 Fix problem with one_time used on a redirect router which returned the
853 parent address unchanged. A retry would see the parent address marked as
854 delivered, so not attempt the (identical) child. As a result mail would
857 JH/27 Fix a possible security hole, wherein a process operating with the Exim
858 UID can gain a root shell. Credit to http://www.halfdog.net/ for
859 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
862 JH/28 Enable {spool,log} filesystem space and inode checks as default.
863 Main config options check_{log,spool}_{inodes,space} are now
864 100 inodes, 10MB unless set otherwise in the configuration.
866 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
867 Previously it only applied to the main-section connection policy
870 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
872 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
873 by me. Added RFC7919 DH primes as an alternative.
875 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
876 are not in the system include path.
878 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
879 GnuTLS, when a session startup failed (eg because the client disconnected)
880 Exim did stdio operations after fclose. This was exposed by a recent
881 change which nulled out the file handle after the fclose.
883 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
884 signed directly by the cert-signing cert, rather than an intermediate
885 OCSP-signing cert. This is the model used by LetsEncrypt.
887 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
889 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
890 an incoming connection.
892 HS/02 Bug 1802: Do not half-close the connection after sending a request
895 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
896 fallback to "prime256v1".
898 JH/34 SECURITY: Use proper copy of DATA command in error message.
899 Could leak key material. Remotely exploitable. CVE-2016-9963.
905 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
906 and 3.4.4 - once the server is enabled to respond to an OCSP request
907 it does even when not requested, resulting in a stapling non-aware
908 client dropping the TLS connection.
910 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
911 support variable-length bit vectors. No functional change.
913 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
914 The I= interface field on outgoing lines is now after the H= remote
915 host field, same as incoming lines. There is a separate
916 outgoing_interface log selector which allows you to disable the
919 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
920 If not running log_selector +smtp_connection the mainlog would be held
921 open indefinitely after a "too many connections" event, including to a
922 deleted file after a log rotate. Leave the per net connection logging
923 leaving it open for efficiency as that will be quickly detected by the
924 check on the next write.
926 HS/01 Bug 1671: Fix post transport crash.
927 Processing the wait-<transport> messages could crash the delivery
928 process if the message IDs didn't exist for some reason. When
929 using 'split_spool_directory=yes' the construction of the spool
930 file name failed already, exposing the same netto behaviour.
932 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
933 mime_regex ACL conditions.
935 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
936 to DSN fail messages (bounces): remote IP, remote greeting, remote response
937 to HELO, local diagnostic string.
939 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
940 log line to debug. Even when configured with a tls authenticator many
941 client connections are expected to not authenticate in this way, so
942 an authenticate fail is not an error.
944 HS/02 Add the Exim version string to the process info. This way exiwhat
945 gives some more detail about the running daemon.
947 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
948 matter for fast-change records such as DNSBLs.
950 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
951 retry record, even if constant. There may be multiple transports with
952 different interface settings and the retry behaviour needs to be kept
955 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
957 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
959 JH/10 Bug 840: fix log_defer_output option of pipe transport
961 JH/11 Bug 830: use same host for all RCPTS of a message, even under
962 hosts_randomize. This matters a lot when combined with mua_wrapper.
964 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
965 ${quote_pgsql:<string>} operator.
967 JH/13 Bug 1708: avoid misaligned access in cached lookup.
969 JH/14 Change header file name for freeradius-client. Relevant if compiling
970 with Radius support; from the Gentoo tree and checked under Fedora.
972 JH/15 Bug 1712: Introduce $prdr_requested flag variable
974 JH/16 Bug 1714: Permit an empty string as expansion result for transport
975 option transport_filter, meaning no filtering.
977 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
979 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
980 defaults to "*" (all hosts). The variable is now available when not built
981 with TLS, default unset, mainly to enable keeping the testsuite sane.
982 If a server certificate is not supplied (via tls_certificate) an error is
983 logged, and clients will find TLS connections fail on startup. Presumably
984 they will retry in-clear.
985 Packagers of Exim are strongly encouraged to create a server certificate
986 at installation time.
988 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
989 with the $config_file variable.
991 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
992 in transport context, after the attempt, and per-recipient. The latter type
993 is per host attempted. The event data is the error message, and the errno
994 information encodes the lookup type (A vs. MX) used for the (first) host,
995 and the trailing two digits of the smtp 4xx response.
997 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
998 to write to mainlog (or rejectlog, paniclog) in the window between file
999 creation and permissions/ownership being changed. Particularly affects
1000 installations where exicyclog is run as root, rather than exim user;
1001 result is that the running daemon panics and dies.
1003 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1005 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1006 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1007 "pri" and "weight". Note that the previous implicit priority given by the
1008 list order is no longer honoured.
1010 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1011 for DKIM processing.
1013 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1014 by defining SUPPORT_SOCKS.
1016 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1017 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1018 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1019 variables are renamed to proxy_{local,external}_{address,port}.
1021 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1022 for a build by defining SUPPORT_I18N
1024 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1025 of the query string, and make ${quote_redis:} do that quoting.
1027 JH/29 Move Events support from Experimental to mainline, enabled by default
1028 and removable for a build by defining DISABLE_EVENT.
1030 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1032 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1033 cached by the daemon.
1035 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1036 by defining LOOKUP_REDIS. The libhiredis library is required.
1038 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1039 keys are given for lookup.
1041 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1042 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1043 only supported when built with TLS support. The PolarSSL SHA routines
1044 are still used when the TLS library is too old for convenient support.
1046 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1047 openssl_options), for security. OpenSSL forces this from version 1.1.0
1048 server-side so match that on older versions.
1050 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1051 allocation for $value could be released as the expansion processing
1052 concluded, but leaving the global pointer active for it.
1054 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1055 and to use the domains and local_parts ACL conditions.
1057 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1058 incorrectly not doubled on cutthrough transmission, hence seen as a
1059 body-termination at the receiving system - resulting in truncated mails.
1060 Commonly the sender saw a TCP-level error, and retransmitted the message
1061 via the normal store-and-forward channel. This could result in duplicates
1062 received - but deduplicating mailstores were liable to retain only the
1063 initial truncated version.
1065 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1067 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1069 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1070 we're in there, support oversigning also; bug 1309.
1072 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1074 HS/04 Add support for keep_environment and add_environment options.
1076 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1077 either intentional arithmetic overflow during PRNG, or testing config-
1080 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1081 delivery resulted in actual delivery. Cancel cutthrough before DATA
1084 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1085 rejects a recipient: pass the reject to the originator.
1087 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1088 Many were false-positives and ignorable, but it's worth fixing the
1091 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1092 for the new environment-manipulation done at startup. Move the routines
1093 from being local to tls.c to being global via the os.c file.
1095 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1096 an extract embedded as result-arg for a map, the first arg for extract
1097 is unavailable so we cannot tell if this is a numbered or keyed
1098 extraction. Accept either.
1104 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1107 JH/02 The smtp transport option "multi_domain" is now expanded.
1109 JH/03 The smtp transport now requests PRDR by default, if the server offers
1112 JH/04 Certificate name checking on server certificates, when exim is a client,
1113 is now done by default. The transport option tls_verify_cert_hostnames
1114 can be used to disable this per-host. The build option
1115 EXPERIMENTAL_CERTNAMES is withdrawn.
1117 JH/05 The value of the tls_verify_certificates smtp transport and main options
1118 default to the word "system" to access the system default CA bundle.
1119 For GnuTLS, only version 3.0.20 or later.
1121 JH/06 Verification of the server certificate for a TLS connection is now tried
1122 (but not required) by default. The verification status is now logged by
1123 default, for both outbound TLS and client-certificate supplying inbound
1126 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1129 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1130 Status Notification (bounce) messages are now MIME format per RFC 3464.
1131 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1132 under the control of the dsn_advertise_hosts option, and routers may
1133 have a dsn_lasthop option.
1135 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1136 default, modifiable by a malware= option. The list separator for
1137 the options can now be changed in the usual way. Bug 68.
1139 JH/10 The smtp_receive_timeout main option is now expanded before use.
1141 JH/11 The incoming_interface log option now also enables logging of the
1142 local interface on delivery outgoing connections.
1144 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1145 if the interface and destination host and port all match.
1147 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1150 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1151 Patch from Andrew Lewis.
1153 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1154 now supports optional time-restrictions, weighting, and priority
1155 modifiers per server. Patch originally by <rommer@active.by>.
1157 JH/16 The spamd_address main option now supports a mixed list of local
1158 and remote servers. Remote servers can be IPv6 addresses, and
1159 specify a port-range.
1161 JH/17 Bug 68: The spamd_address main option now supports an optional
1162 timeout value per server.
1164 JH/18 Bug 1581: Router and transport options headers_add/remove can
1165 now have the list separator specified.
1167 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1170 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1173 JH/21 Support for the A6 type of dns record is withdrawn.
1175 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1176 rather than the verbs used.
1178 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1179 from 255 to 1024 chars.
1181 JH/24 Verification callouts now attempt to use TLS by default.
1183 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1184 are generic router options now. The defaults didn't change.
1186 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1187 Original patch from Alexander Shikoff, worked over by JH.
1189 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1190 files if ZCAT_COMMAND is not executable.
1192 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1194 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1196 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1197 Normally benign, it bites when the pair was led to by a CNAME;
1198 modern usage is to not canonicalize the domain to a CNAME target
1199 (and we were inconsistent anyway for A-only vs AAAA+A).
1201 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1203 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1204 when evaluating $sender_host_dnssec.
1206 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1207 $sender_helo_dnssec variable.
1209 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1211 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1213 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1215 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1216 documented as working, but never had. Support all but $spam_report.
1218 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1219 added for tls authenticator.
1221 HS/03 Add perl_taintmode main config option
1227 TL/01 When running the test suite, the README says that variables such as
1228 no_msglog_check are global and can be placed anywhere in a specific
1229 test's script, however it was observed that placement needed to be near
1230 the beginning for it to behave that way. Changed the runtest perl
1231 script to read through the entire script once to detect and set these
1232 variables, reset to the beginning of the script, and then run through
1233 the script parsing/test process like normal.
1235 TL/02 The BSD's have an arc4random API. One of the functions to induce
1236 adding randomness was arc4random_stir(), but it has been removed in
1237 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1238 function when detected.
1240 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1241 cause callback expansion.
1243 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1244 syntax errors in an expansion can be treated as a string instead of
1245 logging or causing an error, due to the internal use of bool_lax
1246 instead of bool when processing it.
1248 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1249 server certificates when making smtp deliveries.
1251 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1253 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1255 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1257 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1258 Merged patch from Sebastian Wiedenroth.
1260 JH/05 Fix results-pipe from transport process. Several recipients, combined
1261 with certificate use, exposed issues where response data items split
1262 over buffer boundaries were not parsed properly. This eventually
1263 resulted in duplicates being sent. This issue only became common enough
1264 to notice due to the introduction of connection certificate information,
1265 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1267 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1268 size buffer was used, resulting in syntax errors when an expansion
1271 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1272 version 3.3.6 or later.
1274 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1275 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1276 both become "event_action", the variables become $event_name, $event_data
1277 and $event_defer_errno. There is a new variable $verify_mode, usable in
1278 routers, transports and related events. The tls:cert event is now also
1279 raised for inbound connections, if the main configuration event_action
1282 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1283 early OCSP support, but no stapling (appears to be less than 1.0.0).
1285 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1286 server certificate names available under the smtp transport option
1287 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1290 JH/10 Time-related extraction expansions from certificates now use the main
1291 option "timezone" setting for output formatting, and are consistent
1292 between OpenSSL and GnuTLS compilations. Bug 1541.
1294 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1295 encoded parameter in the incoming message. Bug 1558.
1297 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1298 include certificate info, eximon was claiming there were spoolfile
1301 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1303 JH/14 Log delivery-related information more consistently, using the sequence
1304 "H=<name> [<ip>]" wherever possible.
1306 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1307 are problematic for Debian distribution, omit them from the release
1310 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1312 JH/16 Fix string representation of time values on 64bit time_t architectures.
1315 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1316 output list separator was used.
1321 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1322 checkers that were complaining about end of non-void function with no
1325 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1326 This was a regression introduced in 4.83 by another bugfix.
1328 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1330 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1331 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1337 TF/01 Correctly close the server side of TLS when forking for delivery.
1339 When a message was received over SMTP with TLS, Exim failed to clear up
1340 the incoming connection properly after forking off the child process to
1341 deliver the message. In some situations the subsequent outgoing
1342 delivery connection happened to have the same fd number as the incoming
1343 connection previously had. Exim would try to use TLS and fail, logging
1344 a "Bad file descriptor" error.
1346 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1347 utilities have not been installed.
1349 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1350 temporary space as the ACL may create new global variables.
1352 TL/01 LDAP support uses per connection or global context settings, depending
1353 upon the detected version of the libraries at build time.
1355 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1356 to extract and use the src ip:port in logging and expansions as if it
1357 were a direct connection from the outside internet. PPv2 support was
1358 updated based on HAProxy spec change in May 2014.
1360 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1362 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1363 Properly escape header and check for NULL return.
1365 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1368 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1370 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1371 characters in header names, implemented as a verify condition.
1372 Contributed by Michael Fischer v. Mollard.
1374 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1375 results permerror and temperror. Previous values are deprecated but
1376 still accepted. In a future release, err_perm and err_temp will be
1377 completely removed, which will be a backward incompatibility if the
1378 ACL tests for either of these two old results. Patch contributed by
1379 user bes-internal on the mailing list.
1381 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1383 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1384 selectors, in both main and reject logs.
1386 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1389 JH/07 Add malware type "sock" for talking to simple daemon.
1391 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1393 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1394 routers/transports under cutthrough routing.
1396 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1397 numbers. Touch up "bool" conditional to keep the same definition.
1399 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1401 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1403 JH/12 Expand items in router/transport headers_add or headers_remove lists
1404 individually rather than the list as a whole. Bug 1452.
1406 Required for reasonable handling of multiple headers_ options when
1407 they may be empty; requires that headers_remove items with embedded
1408 colons must have them doubled (or the list-separator changed).
1410 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1411 view the policy declared in the DMARC record. Currently, $dmarc_status
1412 is a combined value of both the record presence and the result of the
1415 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1417 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1418 dnslookup router and the smtp transport (applying to the forward
1421 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1422 of ldap servers used for a specific lookup. Patch provided by Heiko
1425 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1426 New variable $lookup_dnssec_authenticated for observability.
1428 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1429 Patch submitted by Lars Timman.
1431 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1433 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1434 Requires trusted mode and valid format message id, aborts otherwise.
1435 Patch contributed by Heiko Schlichting.
1437 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1438 certextract with support for various fields. Bug 1358.
1440 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1441 is requested by default, modifiable by smtp transport option
1444 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1445 operate on certificate variables to give certificate fingerprints
1446 Also new ${sha256:cert_variable}.
1448 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1450 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1451 Christian Aistleitner.
1453 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1455 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1456 file. Patch from Wolfgang Breyha.
1458 JH/25 Expand the coverage of the delivery $host and $host_address to
1459 client authenticators run in verify callout. Bug 1476.
1461 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1462 align with daemon_smtp_ports. Bug 72.
1464 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1465 support and error reporting did not work properly.
1467 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1468 and is readable. Patch from Andrew Colin Kissa.
1470 TL/14 Enhance documentation of ${run expansion and how it parses the
1471 commandline after expansion, particularly in the case when an
1472 unquoted variable expansion results in an empty value.
1474 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1476 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1477 by renaming away from it. Addresses GH issue 15, reported by
1480 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1482 TL/15 SECURITY: prevent double expansion in math comparison functions
1483 (can expand unsanitized data). Not remotely exploitable.
1490 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1492 PP/02 Make -n do something, by making it not do something.
1493 When combined with -bP, the name of an option is not output.
1495 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1498 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1499 $sender_host_name and config options to manage this, and basic check
1502 PP/05 DSCP support for outbound connections and control modifier for inbound.
1504 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1505 (Only plugin which currently uses this is kerberos4, which nobody should
1506 be using, but we should make it available and other future plugins might
1507 conceivably use it, even though it would break NAT; stuff *should* be
1508 using channel bindings instead).
1510 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1511 name; added for Sendmail compatibility; requires admin caller.
1512 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1513 just ignore it); requires trusted caller.
1514 Also parse but ignore: -Ac -Am -X<logfile>
1517 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1519 TL/02 Add +smtp_confirmation as a default logging option.
1521 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1522 Patch by Magnus Holmgren from 2007-02-20.
1524 TL/04 Bugzilla 1281 - Spec typo.
1525 Bugzilla 1283 - Spec typo.
1526 Bugzilla 1290 - Spec grammar fixes.
1528 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1530 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1532 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1533 mailing list by by Dmitry Isaikin.
1535 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1537 JH/02 Support "G" suffix to numbers in ${if comparisons.
1539 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1541 NM/01 Bugzilla 1197 - Spec typo
1542 Bugzilla 1196 - Spec examples corrections
1544 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1546 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1547 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1550 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1551 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1553 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1554 "acl {{name}{arg}...}", and optional args on acl condition
1557 JH/05 Permit multiple router/transport headers_add/remove lines.
1559 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1561 JH/07 Avoid using a waiting database for a single-message-only transport.
1562 Performance patch from Paul Fisher. Bugzilla 1262.
1564 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1567 JH/09 Add $headers_added variable, with content from use of ACL modifier
1568 add_header (but not yet added to the message). Bugzilla 199.
1570 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1571 Pulled from Bugzilla 817 by Wolfgang Breyha.
1573 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1575 (nb: this is the same fix as in Exim 4.80.1)
1577 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1580 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1583 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1585 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1586 advertise SMTP AUTH mechanism to us, instead of a generic
1587 protocol violation error. Also, make Exim more robust to bad
1588 data from the Dovecot auth socket.
1590 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1592 When a queue runner is handling a message, Exim first routes the
1593 recipient addresses, during which it prunes them based on the retry
1594 hints database. After that it attempts to deliver the message to
1595 any remaining recipients. It then updates the hints database using
1598 So if a recipient address works intermittently, it can get repeatedly
1599 deferred at routing time. The retry hints record remains fresh so the
1600 address never reaches the final cutoff time.
1602 This is a fairly common occurrence when a user is bumping up against
1603 their storage quota. Exim had some logic in its local delivery code
1604 to deal with this. However it did not apply to per-recipient defers
1605 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1607 This change adds a proper retry rule check during routing so that the
1608 final cutoff time is checked against the message's age. We only do
1609 this check if there is an address retry record and there is not a
1610 domain retry record; this implies that previous attempts to handle
1611 the address had the retry_use_local_parts option turned on. We use
1612 this as an approximation for the destination being like a local
1613 delivery, as in LMTP.
1615 I suspect this new check makes the old local delivery cutoff check
1616 redundant, but I have not verified this so I left the code in place.
1618 TF/02 Correct gecos expansion when From: is a prefix of the username.
1620 Test 0254 submits a message to Exim with the header
1624 When I ran the test suite under the user fanf2, Exim expanded
1625 the header to contain my full name, whereas it should have added
1626 a Resent-Sender: header. It erroneously treats any prefix of the
1627 username as equal to the username.
1629 This change corrects that bug.
1631 GF/01 DCC debug and logging tidyup
1632 Error conditions log to paniclog rather than rejectlog.
1633 Debug lines prefixed by "DCC: " to remove any ambiguity.
1635 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1637 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1638 Bug spotted by Jeremy Harris; was flawed since initial commit.
1639 Would have resulted in OCSP responses post-SNI triggering an Exim
1640 NULL dereference and crash.
1642 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1644 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1645 Bug detection, analysis and fix by Samuel Thibault.
1646 Bugzilla 1331, Debian bug #698092.
1648 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1650 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1651 Server implementation by Todd Lyons, client by JH.
1652 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1653 config variable "prdr_enable" controls whether the server
1654 advertises the facility. If the client requests PRDR a new
1655 acl_data_smtp_prdr ACL is called once for each recipient, after
1656 the body content is received and before the acl_smtp_data ACL.
1657 The client is controlled by both of: a hosts_try_prdr option
1658 on the smtp transport, and the server advertisement.
1659 Default client logging of deliveries and rejections involving
1660 PRDR are flagged with the string "PRDR".
1662 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1663 fclose(). Diagnosis by Todd Lyons.
1665 PP/17 Update configure.default to handle IPv6 localhost better.
1666 Patch by Alain Williams (plus minor tweaks).
1669 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1670 This is now consistent with GnuTLS, and is now documented: the
1671 previous undocumented portable approach to treating the option as
1672 unset was to force an expansion failure. That still works, and
1673 an empty string is now equivalent.
1675 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1676 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1677 not performing validation itself.
1679 PP/20 Added force_command boolean option to pipe transport.
1680 Patch from Nick Koston, of cPanel Inc.
1682 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1685 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1687 PP/21 Fix eximon continuous updating with timestamped log-files.
1688 Broken in a format-string cleanup in 4.80, missed when I repaired the
1689 other false fix of the same issue.
1690 Report and fix from Heiko Schlichting.
1693 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1694 Report from Prashanth Katuri.
1696 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1697 It's SecureTransport, so affects any MacOS clients which use the
1698 system-integrated TLS libraries, including email clients.
1700 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1701 using a MIME ACL for non-SMTP local injection.
1702 Report and assistance in diagnosis by Warren Baker.
1704 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1706 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1708 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1709 last id that failed so it may be referenced in subsequent ACL's.
1711 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1714 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1715 ldap library initialization, allowing self-signed CA's to be
1716 used. Also properly sets require_cert option later in code by
1717 using NULL (global ldap config) instead of ldap handle (per
1718 session). Bug diagnosis and testing by alxgomz.
1720 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1721 the src/util/ subdirectory.
1723 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1724 renamed to Transport Post Delivery Action by Jeremy Harris, as
1727 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1728 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1729 redis_servers = needs to be configured which will be used by the redis
1730 lookup. Patch from Warren Baker, of The Packet Hub.
1732 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1734 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1735 hostname or reverse DNS when processing a host list. Used suggestions
1736 from multiple comments on this bug.
1738 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1740 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1741 Missed a few lines, added it to make the runtest require no keyboard
1744 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1745 contains upper case chars. Make router use caseful_local_part.
1747 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1748 support when GnuTLS has been built with p11-kit.
1754 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1756 This, or similar/improved, will also be change PP/11 of 4.82.
1762 PP/01 Handle short writes when writing local log-files.
1763 In practice, only affects FreeBSD (8 onwards).
1764 Bugzilla 1053, with thanks to Dmitry Isaikin.
1766 NM/01 Bugzilla 949 - Documentation tweak
1768 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1771 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1773 PP/02 Implemented gsasl authenticator.
1775 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1777 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1778 `pkg-config foo` for cflags/libs.
1780 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1781 with rest of GSASL and with heimdal_gssapi.
1783 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1784 `pkg-config foo` for cflags/libs for the TLS implementation.
1786 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1787 properties get this fed in as external SSF. A number of robustness
1788 and debugging improvements to the cyrus_sasl authenticator.
1790 PP/08 cyrus_sasl server now expands the server_realm option.
1792 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1793 Patch by Jeremy Harris.
1795 PP/10 Added dbmjz lookup type.
1797 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1799 PP/12 MAIL args handles TAB as well as SP, for better interop with
1800 non-compliant senders.
1801 Analysis and variant patch by Todd Lyons.
1803 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1804 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1805 Patch from Dirk Mueller <dmueller@suse.com>
1807 PP/13 tls_peerdn now print-escaped for spool files.
1808 Observed some $tls_peerdn in wild which contained \n, which resulted
1809 in spool file corruption.
1811 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1812 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1813 or write after TLS renegotiation, which otherwise led to messages
1816 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1817 as a tracking header (ie: a signed header comes before the signature).
1818 Patch from Wolfgang Breyha.
1820 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1821 comma-sep list; embedded commas doubled.
1823 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1825 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1827 Report and patch from Dmitry Banschikov.
1829 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1830 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1831 failures. We appear to now support TLS1.1+ with Exim.
1833 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1834 lets Exim select keys and certificates based upon TLS SNI from client.
1835 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1836 before an outbound SMTP session. New log_selector, +tls_sni.
1838 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1839 NULL dereference. Report and patch from Alun Jones.
1841 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1842 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1843 Not seeing resolver debug output on NetBSD, but suspect this is a
1844 resolver implementation change.
1846 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1847 Left warnings. Added "eximon gdb" invocation mode.
1849 PP/21 Defaulting "accept_8bitmime" to true, not false.
1851 PP/22 Added -bw for inetd wait mode support.
1853 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1854 locate the relevant includes and libraries. Made this the default.
1856 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1857 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1859 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1860 This may cause build issues on older platforms.
1862 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1863 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1864 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1865 Added SNI support via GnuTLS too.
1866 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1868 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1870 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1871 Applied second patch from Janne, implementing suggestion to default
1872 multiple-strings-in-record handling to match SPF spec.
1874 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1876 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1877 read-only, out of scope).
1878 Patch from Wolfgang Breyha, report from Stuart Northfield.
1880 PP/29 Fix three issues highlighted by clang analyser static analysis.
1881 Only crash-plausible issue would require the Cambridge-specific
1882 iplookup router and a misconfiguration.
1883 Report from Marcin Mirosław.
1885 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1887 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1888 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1889 As part of this, removing so much warning spew let me fix some minor
1890 real issues in debug logging.
1892 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1893 assignment on my part. Fixed.
1895 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1896 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1897 Janne Snabb (who went above and beyond: thank you).
1899 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1900 string otherwise requires a connection and a bunch more work and it's
1901 relatively easy to get wrong. Should also expose TLS library linkage
1904 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1905 64-bit ${eval} (JH/03).
1907 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1908 GNU libc to support some of the 64-bit stuff, should not lead to
1909 conflicts. Defined before os.h is pulled in, so if a given platform
1910 needs to override this, it can.
1912 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1913 protection layer was required, which is not implemented.
1914 Bugzilla 1254, patch from Wolfgang Breyha.
1916 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1917 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1918 tls_dhparam take prime identifiers. Also unbreak combination of
1919 OpenSSL+DH_params+TLSSNI.
1921 PP/39 Disable SSLv2 by default in OpenSSL support.
1927 PP/01 Solaris build fix for Oracle's LDAP libraries.
1928 Bugzilla 1109, patch from Stephen Usher.
1930 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1932 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1935 TF/02 Fix a couple more cases where we did not log the error message
1936 when unlink() failed. See also change 4.74-TF/03.
1938 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1939 lock up or crash if it happened to be inside a call to libc when it
1940 got a SIGUSR1 from exiwhat.
1942 The SIGUSR1 handler appends the current process status to the process
1943 log which is later printed by exiwhat. It used to use the general
1944 purpose logging code to do this, but several functions it calls are
1945 not safe for signals.
1947 The new output code in the SIGUSR1 handler is specific to the process
1948 log, and simple enough that it's easy to inspect for signal safety.
1949 Removing some special cases also simplifies the general logging code.
1950 Removing the spurious timestamps from the process log simplifies
1953 TF/04 Improved ratelimit ACL condition.
1955 The /noupdate option has been deprecated in favour of /readonly which
1956 has clearer semantics. The /leaky, /strict, and /readonly update modes
1957 are mutually exclusive. The update mode is no longer included in the
1958 database key; it just determines when the database is updated. (This
1959 means that when you upgrade Exim will forget old rate measurements.)
1961 Exim now checks that the per_* options are used with an update mode that
1962 makes sense for the current ACL. For example, when Exim is processing a
1963 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1964 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1965 must specify per_mail/readonly. If you omit the update mode it defaults to
1966 /leaky where that makes sense (as before) or /readonly where required.
1968 The /noupdate option is now undocumented but still supported for
1969 backwards compatibility. It is equivalent to /readonly except that in
1970 ACLs where /readonly is required you may specify /leaky/noupdate or
1971 /strict/noupdate which are treated the same as /readonly.
1973 A useful new feature is the /count= option. This is a generalization
1974 of the per_byte option, so that you can measure the throughput of other
1975 aggregate values. For example, the per_byte option is now equivalent
1976 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1978 The per_rcpt option has been generalized using the /count= mechanism
1979 (though it's more complicated than the per_byte equivalence). When it is
1980 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1981 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1982 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1983 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1984 Note that using per_rcpt with a non-readonly update mode in more than
1985 one ACL will cause the recipients to be double-counted. (The per_mail
1986 and per_byte options don't have this problem.)
1988 The handling of very low rates has changed slightly. If the computed rate
1989 is less than the event's count (usually one) then this event is the first
1990 after a long gap. In this case the rate is set to the same as this event's
1991 count, so that the first message of a spam run is counted properly.
1993 The major new feature is a mechanism for counting the rate of unique
1994 events. The new per_addr option counts the number of different
1995 recipients that someone has sent messages to in the last time period. It
1996 behaves like per_rcpt if all the recipient addresses are different, but
1997 duplicate recipient addresses do not increase the measured rate. Like
1998 the /count= option this is a general mechanism, so the per_addr option
1999 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2000 example, measure the rate that a client uses different sender addresses
2001 with the options per_mail/unique=$sender_address. There are further
2002 details in the main documentation.
2004 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2006 TF/06 Removed a few PCRE remnants.
2008 TF/07 Automatically extract Exim's version number from tags in the git
2009 repository when doing development or release builds.
2011 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2012 Bugzilla 879. Patch from Paul Fisher.
2014 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2015 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2018 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2020 PP/05 Variable $av_failed, true if the AV scanner deferred.
2021 Bugzilla 1078. Patch from John Horne.
2023 PP/06 Stop make process more reliably on build failure.
2024 Bugzilla 1087. Patch from Heiko Schlittermann.
2026 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2027 Bugzilla 1089. Patch from Heiko Schlittermann.
2029 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2030 Bugzilla 1131. Patch from Holger Weiß.
2032 PP/09 Handle IPv6 addresses with SPF.
2033 Bugzilla 860. Patch from Wolfgang Breyha.
2035 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2037 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2040 PP/11 match_* no longer expand right-hand-side by default.
2041 New compile-time build option, EXPAND_LISTMATCH_RHS.
2042 New expansion conditions, "inlist", "inlisti".
2044 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2046 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2048 PP/14 fix log_write() format string regression from TF/03.
2049 Bugzilla 1152. Patch from Dmitry Isaikin.
2055 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2057 PP/02 Harmonised TLS library version reporting; only show if debugging.
2058 Layout now matches that introduced for other libraries in 4.74 PP/03.
2060 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2062 PP/04 New "dns_use_edns0" global option.
2064 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2067 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2068 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2070 TK/01 Updated PolarSSL code to 0.14.2.
2071 Bugzilla 1097. Patch from Andreas Metzler.
2073 PP/07 Catch divide-by-zero in ${eval:...}.
2074 Fixes bugzilla 1102.
2076 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2079 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2080 format-string attack -- SECURITY: remote arbitrary code execution.
2082 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2083 time unintentionally subject to list matching rules, letting the header
2084 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2085 arbitrary string expansion). This allowed for information disclosure.
2087 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2088 INT_MIN/-1 -- value coerced to INT_MAX.
2094 NM/01 Workaround for PCRE version dependency in version reporting
2097 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2098 This fixes portability to compilers other than gcc, notably
2099 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2101 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2102 makefiles for portability to HP-UX and POSIX correctness.
2104 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2105 Also via indented variable definition in the Makefile.
2106 (Debugging by Oliver Heesakkers).
2108 PP/02 Restore caching of spamd results with expanded spamd_address.
2109 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2111 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2112 Improves build reliability. Fix from: Frank Elsner
2114 NM/02 Fix wide character breakage in the rfc2047 coding
2115 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2117 NM/03 Allow underscore in dnslist lookups
2118 Fixes bug 1026. Patch from Graeme Fowler
2120 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2121 Code patches from Adam Ciarcinski of NetBSD.
2123 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2126 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2127 is logged, to avoid truncation. Patch from John Horne.
2129 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2130 Patch from Jakob Hirsch.
2132 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2133 SQL string expansion failure details.
2134 Patch from Andrey Oktyabrski.
2136 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2137 Patch from Simon Arlott.
2139 PP/09 New lookups functionality failed to compile on old gcc which rejects
2140 extern declarations in function scope.
2141 Patch from Oliver Fleischmann
2143 PP/10 Use sig_atomic_t for flags set from signal handlers.
2144 Check getgroups() return and improve debugging.
2145 Fixed developed for diagnosis in bug 927 (which turned out to be
2148 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2149 Patch from Mark Zealey.
2151 PP/12 Bugzilla 1056: Improved spamd server selection.
2152 Patch from Mark Zealey.
2154 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2155 Based on patch from Heiko Schlittermann.
2157 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2158 Patch from Uwe Doering, sign-off by Michael Haardt.
2160 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2161 variable declaration deep within a block. Bug and patch from
2164 PP/15 lookups-Makefile IRIX compatibility coercion.
2166 PP/16 Make DISABLE_DKIM build knob functional.
2168 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2169 Patch by Simon Arlott
2171 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2172 variable argument macros. Our copy now differs from upstream.
2178 TF/01 Failure to get a lock on a hints database can have serious
2179 consequences so log it to the panic log.
2181 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2182 controlled using the smtp_confirmation log selector.
2184 TF/03 Include the error message when we fail to unlink a spool file.
2186 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2187 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2188 for maintaining out-of-tree patches for some time.
2190 PP/01 Bugzilla 139: Documentation and portability issues.
2191 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2192 Handle per-OS dynamic-module compilation flags.
2194 PP/02 Let /dev/null have normal permissions.
2195 The 4.73 fixes were a little too stringent and complained about the
2196 permissions on /dev/null. Exempt it from some checks.
2197 Reported by Andreas M. Kirchwitz.
2199 PP/03 Report version information for many libraries, including
2200 Exim version information for dynamically loaded libraries. Created
2201 version.h, now support a version extension string for distributors
2202 who patch heavily. Dynamic module ABI change.
2204 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2205 privilege escalation vulnerability whereby the Exim run-time user
2206 can cause root to append content of the attacker's choosing to
2209 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2212 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2213 If dropping privileges for untrusted macros, we disabled normal logging
2214 on the basis that it would fail; for the Exim run-time user, this is not
2215 the case, and it resulted in successful deliveries going unlogged.
2216 Fixed. Reported by Andreas Metzler.
2222 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2223 only prepend for the Resent-* case. Fixes regression introduced in
2224 Exim 4.70 by NM/22 for Bugzilla 607.
2226 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2227 increasing numbers of administrators be bitten by this.
2229 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2231 PP/03 Bugzilla 994: added openssl_options main configuration option.
2233 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2235 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2237 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2239 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2240 without a peer certificate, leading to a segfault because of an
2241 assumption that peers always have certificates. Be a little more
2242 paranoid. Problem reported by Martin Tscholak.
2244 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2245 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2246 NB: ClamAV planning to remove STREAM in "middle of 2010".
2247 CL also introduces -bmalware, various -d+acl logging additions and
2248 more caution in buffer sizes.
2250 PP/09 Implemented reverse_ip expansion operator.
2252 PP/10 Bugzilla 937: provide a "debug" ACL control.
2254 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2256 PP/12 Bugzilla 973: Implement --version.
2258 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2260 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2262 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2264 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2265 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2266 ignore trailing whitespace.
2268 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2270 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2273 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2274 Notification from Dr Andrew Aitchison.
2276 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2277 ExtendedDetectionInfo response format.
2278 Notification from John Horne.
2280 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2283 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2284 XSL and documented dependency on system catalogs, with examples of how
2287 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2290 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2291 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2292 configuration file which is writeable by the Exim user or group.
2294 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2295 of configuration files to cover files specified with the -C option if
2296 they are going to be used with root privileges, not just the default
2299 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2300 option (effectively making it always true).
2302 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2303 files to be used while preserving root privileges.
2305 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2306 that rogue child processes cannot use them.
2308 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2309 run-time user, instead of root.
2311 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2312 Exim run-time user without dropping privileges.
2314 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2315 result string, instead of calling string_vformat() twice with the same
2318 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2319 for other users. Others should always drop root privileges if they use
2320 -C on the command line, even for a whitelisted configure file.
2322 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2324 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2330 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2331 $data_path, and $header_path variables; fixed documentation bugs and
2334 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2335 exipick to access non-standard spools, including the "frozen" queue
2338 NM/01 Bugzilla 965: Support mysql stored procedures.
2339 Patch from Alain Williams
2341 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2343 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2344 Patch from Andreas Metzler
2346 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2347 Patch from Kirill Miazine
2349 NM/05 Bugzilla 671: Added umask to procmail example.
2351 JJ/03 installed exipick 20100323.0, fixing doc bug
2353 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2354 directory. Notification and patch from Dan Rosenberg.
2356 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2358 TK/02 Improve log output when DKIM signing operation fails.
2360 MH/01 Treat the transport option dkim_domain as a colon separated
2361 list, not as a single string, and sign the message with each element,
2362 omitting multiple occurences of the same signer.
2364 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2365 Bugzilla 985, 986. Patch by Simon Arlott
2367 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2368 Patch by Simon Arlott
2370 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2371 MBX locking. Notification from Dan Rosenberg.
2377 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2379 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2381 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2383 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2385 NM/04 Bugzilla 915: Flag broken perl installation during build.
2391 TK/01 Added patch by Johannes Berg that expands the main option
2392 "spamd_address" if it starts with a dollar sign.
2394 TK/02 Write list of recipients to X-Envelope-Sender header when building
2395 the mbox-format spool file for content scanning (suggested by Jakob
2398 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2399 (http://www.dcc-servers.net/) support via dccifd. Activated by
2400 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2402 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2403 by Mark Daniel Reidel <mr@df.eu>.
2405 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2406 When building exim an external PCRE library is now needed -
2407 PCRE is a system library on the majority of modern systems.
2408 See entry on PCRE_LIBS in EDITME file.
2410 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2411 conversation. Added nologin parameter to request.
2412 Patch contributed by Kirill Miazine.
2414 TF/01 Do not log submission mode rewrites if they do not change the address.
2416 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2418 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2419 log files in place. Contributed by Roberto Lima.
2421 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2423 TF/03 Bugzilla 615: When checking the local_parts router precondition
2424 after a local_part_suffix or local_part_prefix option, Exim now
2425 does not use the address's named list lookup cache, since this
2426 contains cached lookups for the whole local part.
2428 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2429 Robert Millan. Documentation is in experimental-spec.txt.
2431 TF/04 Bugzilla 668: Fix parallel build (make -j).
2433 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2435 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2436 Patch provided by Jan Srzednicki.
2438 TF/05 Leading white space used to be stripped from $spam_report which
2439 wrecked the formatting. Now it is preserved.
2441 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2442 that they are available at delivery time.
2444 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2446 TF/08 TLS error reporting now respects the incoming_interface and
2447 incoming_port log selectors.
2449 TF/09 Produce a more useful error message if an SMTP transport's hosts
2450 setting expands to an empty string.
2452 NM/06 Bugzilla 744: EXPN did not work under TLS.
2453 Patch provided by Phil Pennock.
2455 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2456 Patch provided by Richard Godbee.
2458 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2459 acl_smtp_notquit, added index entry.
2461 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2462 Patch provided by Eugene Bujak.
2464 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2465 accept(). Patch provided by Maxim Dounin.
2467 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2468 Patch provided by Phil Pennock.
2470 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2472 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2473 Patch provided by Brad "anomie" Jorsch.
2475 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2476 Patch provided by Dean Brooks.
2478 TK/05 Add native DKIM support (does not depend on external libraries).
2480 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2481 Patch provided by Graeme Fowler.
2483 NM/16 Bugzilla 851: Documentation example syntax fix.
2485 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2487 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2490 NM/19 Bugzilla 745: TLS version reporting.
2491 Patch provided by Phil Pennock.
2493 NM/20 Bugzilla 167: bool: condition support.
2494 Patch provided by Phil Pennock.
2496 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2497 clients. Patch provided by Phil Pennock.
2499 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2500 Patch provided by Brad "anomie" Jorsch.
2502 NM/23 Bugzilla 687: Fix misparses in eximstats.
2503 Patch provided by Heiko Schlittermann.
2505 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2506 Patch provided by Heiko Schlittermann.
2508 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2509 plus update to original patch.
2511 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2513 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2514 Patch provided by David Brownlee.
2516 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2518 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2520 NM/30 Bugzilla 888: TLS documentation bugfixes.
2522 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2524 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2525 Unlike the original bugzilla I have changed all shell scripts in src tree.
2527 NM/33 Bugzilla 898: Transport filter timeout fix.
2528 Patch by Todd Rinaldo.
2530 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2531 Patch by Serge Demonchaux.
2533 NM/35 Bugzilla 39: Base64 decode bug fixes.
2534 Patch by Jakob Hirsch.
2536 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2538 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2540 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2542 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2548 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2549 ALT-N's libdkim that I have put here:
2550 http://duncanthrax.net/exim-experimental/
2552 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2553 were called 'true' and it seems that C99 defines that as a reserved
2554 keyword to be used with 'bool' variable types. That means you could
2555 not include C99-style headers which use bools without triggering
2556 build errors in sieve.c.
2558 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2559 as mailq or other aliases. Changed the --help handling significantly
2560 to do whats expected. exim_usage() emits usage/help information.
2562 SC/01 Added the -bylocaldomain option to eximstats.
2564 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2566 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2568 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2574 PH/01 Another patch from the Sieve maintainer.
2576 PH/02 When an IPv6 address is converted to a string for single-key lookup
2577 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2578 dots are used instead of colons so that keys in lsearch files need not
2579 contain colons. This was done some time before quoting was made available
2580 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2581 (notated using the quote facility) so as to distinguish them from IPv4
2582 keys. This meant that lookups for IP addresses in host lists did not work
2583 for iplsearch lookups.
2585 This has been fixed by arranging for IPv6 addresses to be expressed with
2586 colons if the lookup type is iplsearch. This is not incompatible, because
2587 previously such lookups could never work.
2589 The situation is now rather anomalous, since one *can* have colons in
2590 ordinary lsearch keys. However, making the change in all cases is
2591 incompatible and would probably break a number of configurations.
2593 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2596 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2597 conversion specification without a maximum field width, thereby enabling
2598 a rogue spamd server to cause a buffer overflow. While nobody in their
2599 right mind would setup Exim to query an untrusted spamd server, an
2600 attacker that gains access to a server running spamd could potentially
2601 exploit this vulnerability to run arbitrary code as the Exim user.
2603 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2604 $primary_hostname instead of what libspf2 thinks the hosts name is.
2606 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2607 a directory entry by the name of the lookup key. Previously, if a
2608 symlink pointed to a non-existing file or a file in a directory that
2609 Exim lacked permissions to read, a lookup for a key matching that
2610 symlink would fail. Now it is enough that a matching directory entry
2611 exists, symlink or not. (Bugzilla 503.)
2613 PH/03 The body_linecount and body_zerocount variables are now exported in the
2616 PH/04 Added the $dnslist_matched variable.
2618 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2619 This means they are set thereafter only if the connection becomes
2622 PH/06 Added the client_condition to authenticators so that some can be skipped
2623 by clients under certain conditions.
2625 PH/07 The error message for a badly-placed control=no_multiline_responses left
2626 "_responses" off the end of the name.
2628 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2630 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2631 (without spaces) instead of just copying the configuration text.
2633 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2635 PH/11 Added $max_received_linelength.
2637 PH/12 Added +ignore_defer and +include_defer to host lists.
2639 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2640 way in which PCRE > 7.0 is built.
2642 PH/14 Implemented queue_only_load_latch.
2644 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2645 MAIL command. The effect was to mangle the value on 64-bit systems.
2647 PH/16 Another patch from the Sieve maintainer.
2649 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2651 PH/18 If a system quota error occurred while trying to create the file for
2652 a maildir delivery, the message "Mailbox is full" was not appended to the
2653 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2654 only to a quota excession during the actual writing of the file.
2656 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2657 characters?) which causes problems in log lines. The DN values are now
2658 passed through string_printing() before being added to log lines.
2660 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2661 and InterBase are left for another time.)
2663 PH/21 Added message_body_newlines option.
2665 PH/22 Guard against possible overflow in moan_check_errorcopy().
2667 PH/23 POSIX allows open() to be a macro; guard against that.
2669 PH/24 If the recipient of an error message contained an @ in the local part
2670 (suitably quoted, of course), incorrect values were put in $domain and
2671 $local_part during the evaluation of errors_copy.
2677 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2678 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2681 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2682 issue a MAIL command.
2684 PH/02 In an ACL statement such as
2686 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2688 if a client was not listed at all, or was listed with a value other than
2689 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2690 the condition was not true (as it should be), so access was not denied.
2691 The bug was that the ! inversion was incorrectly passed on to the second
2692 item. This has been fixed.
2694 PH/03 Added additional dnslists conditions == and =& which are different from
2695 = and & when the dns lookup returns more than one IP address.
2697 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2698 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2700 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2701 FSYNC, which compiles an option called disable_fsync that allows for
2702 bypassing fsync(). The documentation is heavily laced with warnings.
2704 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2706 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2707 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2708 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2709 including adding "make clean"; (3) Added -fPIC when compiling the test
2710 dynamically loaded module, to get rid of a warning.
2712 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2713 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2714 = 0s. The bug is otherwise harmless.
2716 PH/07 There was a bug in the dovecot authenticator such that the value of
2717 $auth1 could be overwritten, and so not correctly preserved, after a
2718 successful authentication. This usually meant that the value preserved by
2719 the server_setid option was incorrect.
2721 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2723 PH/09 Installed PCRE release 7.0.
2725 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2726 run for batched SMTP input. It is now run at the start of every message
2727 in the batch. While fixing this I discovered that the process information
2728 (output by running exiwhat) was not always getting set for -bs and -bS
2729 input. This is fixed, and it now also says "batched" for BSMTP.
2731 PH/11 Added control=no_pipelining.
2733 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2734 patch, slightly modified), and move the expansion of helo_data till after
2735 the connection is made in the smtp transport (so it can use these
2738 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2740 PH/14 Added log_selector = +pid.
2742 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2744 PH/16 Add ${if forany and ${if forall.
2746 PH/17 Added dsn_from option to vary the From: line in DSNs.
2748 PH/18 Flush SMTP output before performing a callout, unless control =
2749 no_callout_flush is set.
2751 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2752 was true (the default) a successful delivery failed to delete the retry
2753 item, thus causing premature timeout of the address. The bug is now
2756 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2758 PH/21 Long custom messages for fakedefer and fakereject are now split up
2759 into multiline responses in the same way that messages for "deny" and
2760 other ACL rejections are.
2762 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2763 with slight modification.
2765 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2766 draft, changing the syntax and factoring some duplicate code".
2768 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2769 for deliveries of the second and subsequent messages over the same SMTP
2772 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2773 ${reduce, with only minor "tidies".
2775 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2777 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2778 expansion side effects.
2780 PH/27 When a message times out after an over-quota error from an Exim-imposed
2781 quota, the bounce message says "mailbox is full". This message was not
2782 being given when it was a system quota that was exceeded. It now should
2785 MH/03 Made $recipients available in local_scan(). local_scan() already has
2786 better access to the recipient list through recipients_list[], but
2787 $recipients can be useful in postmaster-provided expansion strings.
2789 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2790 in the case of a MAIL command with additional options following the
2791 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2792 were accidentally chopped off.
2794 PH/29 SMTP synchronization checks are implemented when a command is read -
2795 there is a check that no more input is waiting when there shouldn't be
2796 any. However, for some commands, a delay in an ACL can mean that it is
2797 some time before the response is written. In this time, more input might
2798 arrive, invalidly. So now there are extra checks after an ACL has run for
2799 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2800 pipelining has not been advertised.
2802 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2804 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2805 returned in the SMTP response when smtp_return_error_details was false.
2806 This has been fixed.
2808 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2809 instead of the C I/O that was originally supplied, because problems were
2810 reported on Solaris.
2812 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2813 Exim which did not show up earlier: it was assuming that a call to
2814 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2815 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2816 was a macro that became an assignment, so it seemed to work. This has
2817 changed to a proper function call with a void return, hence the compile
2818 error. Exim's code has been fixed.
2820 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2823 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2825 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2827 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2828 because it turns out to be too big or there is a timeout) there is no
2829 "Completed" line in the log. When some messages of this type were
2830 selected by exigrep, they were listed as "not completed". Others were
2831 picked up by some special patterns. I have improved the selection
2832 criteria to be more general.
2834 PH/38 The host_find_failed option in the manualroute router can now be set
2835 to "ignore", to completely ignore a host whose IP address cannot be
2836 found. If all hosts are ignored, the behaviour is controlled by the new
2837 host_all_ignored option.
2839 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2840 homing or because of multiple MX records with /mx) generated more than
2841 one IP address, and the following item turned out to be the local host,
2842 all the secondary addresses of the first item were incorrectly removed
2843 from the list, along with the local host and any following hosts (which
2844 is what is supposed to happen).
2846 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2847 whoever called Exim into the -H file. In the case of the daemon it was
2848 behaving confusingly. When first started, it used values for whoever
2849 started the daemon, but after a SIGHUP it used the Exim user (because it
2850 calls itself on a restart). I have changed the code so that it now always
2853 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2854 message are rejected with the same error (e.g. no authentication or bad
2855 sender address), and a DATA command is nevertheless sent (as can happen
2856 with PIPELINING or a stupid MUA), the error message that was given to the
2857 RCPT commands is included in the rejection of the DATA command. This is
2858 intended to be helpful for MUAs that show only the final error to their
2861 PH/42 Another patch from the Sieve maintainer.
2863 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2864 Eximstats - Fixed some broken HTML links and added missing column headers
2866 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2867 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2869 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2871 PH/43 Yet another patch from the Sieve maintainer.
2873 PH/44 I found a way to check for a TCP/IP connection going away before sending
2874 the response to the final '.' that terminates a message, but only in the
2875 case where the client has not sent further data following the '.'
2876 (unfortunately, this is allowed). However, in many cases there won't be
2877 any further data because there won't be any more messages to send. A call
2878 to select() can be used: if it shows that the input is "ready", there is
2879 either input waiting, or the socket has been closed. An attempt to read
2880 the next input character can distinguish the two cases. Previously, Exim
2881 would have sent an OK response which the client would never have see.
2882 This could lead to message repetition. This fix should cure that, at
2883 least in a lot of common cases.
2885 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2886 advertised in response to EHLO.
2892 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2893 fixed by 4.65/MH/01 (is this a record?) are fixed:
2895 (i) An empty string was always treated as zero by the numeric comparison
2896 operators. This behaviour has been restored.
2898 (ii) It is documented that the numeric comparison operators always treat
2899 their arguments as decimal numbers. This was broken in that numbers
2900 starting with 0 were being interpreted as octal.
2902 While fixing these problems I realized that there was another issue that
2903 hadn't been noticed. Values of message_size_limit (both the global option
2904 and the transport option) were treated as octal if they started with 0.
2905 The documentation was vague. These values are now always treated as
2906 decimal, and I will make that clear in the documentation.
2912 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2913 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2916 MH/01 Don't check that the operands of numeric comparison operators are
2917 integers when their expansion is in "skipping" mode (fixes bug
2918 introduced by 4.64-PH/07).
2920 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2921 child addresses, Exim now panics and dies. Previously, because the count
2922 is held in a short int, deliveries were likely to be lost. As such a
2923 large number of recipients for a single message is ridiculous
2924 (performance will be very, very poor), I have chosen to impose a limit
2925 rather than extend the field.
2931 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2932 leftover -K file (the existence of which was triggered by #402).
2933 While we were at it, introduced process PID as part of the -K
2934 filename. This should rule out race conditions when creating
2937 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2938 processing considerably. Previous code took too long for large mails,
2939 triggering a timeout which in turn triggers #401.
2941 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2942 in the DK code in transports.c. sendfile() is not really portable,
2943 hence the _LINUX specificness.
2945 TF/01 In the add_headers option to the mail command in an Exim filter,
2946 there was a bug that Exim would claim a syntax error in any
2947 header after the first one which had an odd number of characters
2950 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2951 callout verification, Exim cached a "reject" for the entire domain. This
2952 is correct for most verifications, but it is not correct for a recipient
2953 verification with use_sender or use_postmaster set, because in that case
2954 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2955 case of MAIL FROM:<> rejection from other early rejections (e.g.
2956 rejection of HELO). When verifying a recipient using a non-null MAIL
2957 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2958 Whatever the result of the callout, the value of the domain cache is
2959 left unchanged (for any other kind of callout, getting as far as trying
2960 RCPT means that the domain itself is ok).
2962 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2965 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2966 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2967 session, the daemon ignores ECONNECT errors and logs others; it now
2968 ignores EPIPE as well.
2970 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2971 (quoted-printable decoding).
2973 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2974 later the small subsequent patch to fix an introduced bug.
2976 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2978 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2980 PH/08 An error is now given if message_size_limit is specified negative.
2982 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2983 to be given (somewhat) arbitrary names.
2985 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2988 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2989 miscellaneous code fixes
2991 PH/10 Added the log_reject_target ACL modifier to specify where to log
2994 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2995 hostname. This is wrong, because it relates to the incoming message (and
2996 probably the interface on which it is arriving) and not to the outgoing
2997 callout (which could be using a different interface). This has been
2998 changed to use the value of the helo_data option from the smtp transport
2999 instead - this is what is used when a message is actually being sent. If
3000 there is no remote transport (possible with a router that sets up host
3001 addresses), $smtp_active_hostname is used.
3003 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3004 tweaks were necessary in order to get it to work (see also 21 below):
3005 (a) The code assumed that strncpy() returns a negative number on buffer
3006 overflow, which isn't the case. Replaced with Exim's string_format()
3008 (b) There were several signed/unsigned issues. I just did the minimum
3009 hacking in of casts. There is scope for a larger refactoring.
3010 (c) The code used strcasecmp() which is not a standard C function.
3011 Replaced with Exim's strcmpic() function.
3012 (d) The code set only $1; it now sets $auth1 as well.
3013 (e) A simple test gave the error "authentication client didn't specify
3014 service in request". It would seem that Dovecot has changed its
3015 interface. Fortunately there's a specification; I followed it and
3016 changed what the client sends and it appears to be working now.
3018 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3021 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3022 address A is aliased to B and C, where B exists and C does not. Without
3023 -v the output is "A verified" because verification stops after a
3024 successful redirection if more than one address is generated. However,
3025 with -v the child addresses are also verified. Exim was outputting "A
3026 failed to verify" and then showing the successful verification for C,
3027 with its parentage. It now outputs "B failed to verify", showing B's
3028 parentage before showing the successful verification of C.
3030 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3031 look up a TXT record in a specific list after matching in a combined
3034 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3035 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3036 they consult the DNS. I had assumed they would set it the way they
3037 wanted; and indeed my experiments on Linux seem to show that in some
3038 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3039 To be on the safe side, however, I have now made the interface to
3040 host_find_byname() similar to host_find_bydns(), with an argument
3041 containing the DNS resolver options. The host_find_byname() function now
3042 sets these options at its start, just as host_find_bydns() does. The smtp
3043 transport options dns_qualify_single and dns_search_parents are passed to
3044 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3045 of host_find_byname() use the default settings of RES_DEFNAMES
3046 (qualify_single) but not RES_DNSRCH (search_parents).
3048 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3049 spool_read_header() do less string testing, by means of a preliminary
3050 switch on the second character of optional "-foo" lines. (This is
3051 overdue, caused by the large number of possibilities that now exist.
3052 Originally there were few.) While I was there, I also converted the
3053 str(n)cmp tests so they don't re-test the leading "-" and the first
3054 character, in the hope this might squeeze out yet more improvement.
3056 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3057 flag allowing group syntax was set by the header_syntax check but not
3058 turned off, possible causing trouble later; (2) The flag was not being
3059 set at all for the header_verify test, causing "group"-style headers to
3060 be rejected. I have now set it in this case, and also caused header_
3061 verify to ignore an empty address taken from a group. While doing this, I
3062 came across some other cases where the code for allowing group syntax
3063 while scanning a header line wasn't quite right (mostly, not resetting
3064 the flag correctly in the right place). These bugs could have caused
3065 trouble for malformed header lines. I hope it is now all correct.
3067 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3068 with the "reply" argument non-NULL. The code, however (which originally
3069 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3070 but it didn't always do it. This confused somebody who was copying the
3071 code for some other use. I have removed all the tests.
3073 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3074 feature that was used to support insecure browsers during the U.S. crypto
3075 embargo. It requires special client support, and Exim is probably the
3076 only MTA that supported it -- and would never use it because real RSA is
3077 always available. This code has been removed, because it had the bad
3078 effect of slowing Exim down by computing (never used) parameters for the
3079 RSA_EXPORT functionality.
3081 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3082 authenticator to fail if there's a tab character in the incoming data
3083 (there should never be unless someone is messing about, as it's supposed
3084 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3085 if the connection is using TLS or if the remote IP is the same as the
3086 local IP, and the "valid-client-cert option" if a client certificate has
3089 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3090 authenticators. This can be used for authorization after authentication
3091 succeeds. (In the case of plaintext, it servers for both authentication
3094 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3095 if any retry times were supplied.
3097 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3098 connection, where there is no client IP address to check. In this
3099 situation, the verify now always succeeds.
3101 PH/25 Applied John Jetmore's -Mset patch.
3103 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3105 PH/27 In a string expansion for a processed (not raw) header when multiple
3106 headers of the same name were present, leading whitespace was being
3107 removed from all of them, but trailing whitespace was being removed only
3108 from the last one. Now trailing whitespace is removed from each header
3109 before concatenation. Completely empty headers in a concatenation (as
3110 before) are ignored.
3112 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3113 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3115 PH/29 [Removed. This was a change that I later backed out, and forgot to
3116 correct the ChangeLog entry (that I had efficiently created) before
3117 committing the later change.]
3119 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3120 address errors (4xx response to RCPT) over the same connection as other
3121 messages routed to the same hosts. Such deliveries are always "forced",
3122 so retry times are not inspected. This resulted in far too many retries
3123 for the affected addresses. The effect occurred only when there were more
3124 hosts than the hosts_max_try setting in the smtp transport when it had
3125 the 4xx errors. Those hosts that it had tried were not added to the list
3126 of hosts for which the message was waiting, so if all were tried, there
3127 was no problem. Two fixes have been applied:
3129 (i) If there are any address or message errors in an SMTP delivery, none
3130 of the hosts (tried or untried) are now added to the list of hosts
3131 for which the message is waiting, so the message should not be a
3132 candidate for sending over the same connection that was used for a
3133 successful delivery of some other message. This seems entirely
3134 reasonable: after all the message is NOT "waiting for some host".
3135 This is so "obvious" that I'm not sure why it wasn't done
3136 previously. Hope I haven't missed anything, but it can't do any
3137 harm, as the worst effect is to miss an optimization.
3139 (ii) If, despite (i), such a delivery is accidentally attempted, the
3140 routing retry time is respected, so at least it doesn't keep
3141 hammering the server.
3143 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3144 in ${readsocket because some servers need this prod.
3146 PH/32 Added some extra debug output when updating a wait-xxx database.
3148 PH/33 The hint "could be header name not terminated by colon", which has been
3149 given for certain expansion errors for a long time, was not being given
3150 for the ${if def:h_colon_omitted{... case.
3152 PH/34 The spec says: "With one important exception, whenever a domain list is
3153 being scanned, $domain contains the subject domain." There was at least
3154 one case where this was not true.
3156 PH/35 The error "getsockname() failed: connection reset by peer" was being
3157 written to the panic log as well as the main log, but it isn't really
3158 panic-worthy as it just means the connection died rather early on. I have
3159 removed the panic log writing for the ECONNRESET error when getsockname()
3162 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3163 runs only) independently of the message's sender address. This meant
3164 that, if the 4xx error was in fact related to the sender, a different
3165 message to the same recipient with a different sender could confuse
3166 things. In particular, this can happen when sending to a greylisting
3167 server, but other circumstances could also provoke similar problems.
3168 I have changed the default so that the retry time for these errors is now
3169 based a combination of the sender and recipient addresses. This change
3170 can be overridden by setting address_retry_include_sender=false in the
3173 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3174 remote server are returned as part of bounce messages. This was not
3175 happening for LMTP over a pipe (the lmtp transport), but now it is the
3176 same for both kinds of LMTP.
3178 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3179 in header lines that were in fact CNAMEs. This is no longer the case.
3181 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3182 and queue runs started by the daemon processed all messages. This has
3183 been fixed so that -R and -S can now usefully be given with -q<time>.
3185 PH/40 Import PCRE release 6.7 (fixes some bugs).
3187 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3189 PH/42 Give an error if -q is specified more than once.
3191 PH/43 Renamed the variables $interface_address and $interface_port as
3192 $received_ip_address and $received_port, to make it clear that these
3193 values apply to message reception, and not to the outgoing interface when
3194 a message is delivered. (The old names remain recognized, of course.)
3196 PH/44 There was no timeout on the connect() call when using a Unix domain
3197 socket in the ${readsocket expansion. There now is.
3199 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3200 be meaningful with "accept".
3202 SC/01 Eximstats V1.43
3203 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3205 SC/02 Eximstats V1.44
3206 Use a glob alias rather than an array ref in the generated
3207 parser. This improves both readability and performance.
3209 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3210 Collect SpamAssassin and rejection statistics.
3211 Don't display local sender or destination tables unless
3212 there is data to show.
3213 Added average volumes into the top table text output.
3215 SC/04 Eximstats V1.46
3216 Collect data on the number of addresses (recipients)
3217 as well as the number of messages.
3219 SC/05 Eximstats V1.47
3220 Added 'Message too big' to the list of mail rejection
3221 reasons (thanks to Marco Gaiarin).
3223 SC/06 Eximstats V1.48
3224 Mainlog lines which have GMT offsets and are too short to
3225 have a flag are now skipped.
3227 SC/07 Eximstats V1.49 (Alain Williams)
3228 Added the -emptyok flag.
3230 SC/08 Eximstats V1.50
3231 Fixes for obtaining the IP address from reject messages.
3233 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3234 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3235 whitespace changes from 4.64-PH/27
3237 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3240 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3241 are found, allow negative numbers in numeric criteria)
3243 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3245 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3246 to match changes made in 4.64-PH/43
3248 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3250 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3251 "domain" is actually an IP address, and force a failure. This locks out
3252 those revolvers/nameservers that support "A-for-A" lookups, in
3253 contravention of the specifications.
3255 PH/48 When a host name was looked up from an IP address, and the subsequent
3256 forward lookup of the name timed out, the host name was left in
3257 $sender_host_name, contrary to the specification.
3259 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3260 restricted to single-key lookups, Exim was not diagnosing an error if
3261 * or *@ was used with a query-style lookup.
3263 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3265 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3266 long ago, but noone interested enough thought of it. Let's just say that
3267 the "1.1" means that there are some new functions that weren't there at
3268 some point in the past.
3270 PH/51 Error processing for expansion failure of helo_data from an smtp
3271 transport during callout processing was broken.
3273 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3274 tested/used via the -bh/-bhc/-bs options.
3276 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3277 bug, fixed in subsequent PCRE releases).
3279 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3280 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3282 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3288 SC/01 Use a glob alias rather than an array ref in eximstats generated
3289 parser. This improves both readability and performance.
3291 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3292 Don't display local sender or destination tables in eximstats unless
3293 there is data to show.
3294 Added average volumes into the eximstats top table text output.
3296 SC/03 Collect data on the number of addresses (recipients) as well
3297 as the number of messages in eximstats.
3299 TF/01 Correct an error in the documentation for the redirect router. Exim
3300 does (usually) call initgroups() when daemonizing.
3302 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3303 with consistent privilege compared to when running as a daemon.
3305 TF/03 Note in the spec that $authenticated_id is not set for local
3306 submissions from trusted users.
3308 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3309 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3311 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3312 by adding some example configuration directives to the default
3313 configuration file. A little bit of work is required to uncomment the
3314 directives and define how usernames and passwords are checked, but
3315 there is now a framework to start from.
3317 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3318 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3319 without this. I don't know how relevant this is to other LDAP libraries.
3321 PH/02 Add the verb name to the "unknown ACL verb" error.
3323 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3325 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3327 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3328 directory not expanded when it should be if an expanded home directory
3329 was set for the address (which is overridden by the transport).
3331 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3334 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3335 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3336 because it is too late at that time, and has no effect.
3338 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3339 security issue with \' (bugzilla #107). I could not use the
3340 PQescapeStringConn() function, because it needs a PGconn value as one of
3343 PH/08 When testing addresses using -bt, indicate those final addresses that
3344 are duplicates that would not cause an additional delivery. At least one
3345 person was confused, thinking that -bt output corresponded to deliveries.
3346 (Suppressing duplicates isn't a good idea as you lose the information
3347 about possibly different redirections that led to the duplicates.)
3349 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3350 systems where poll() doesn't work, in particular OS X.
3352 PH/10 Added more information to debugging output for retry time not reached.
3354 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3355 operations in malware.c.
3357 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3360 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3361 syslog_duplication set false, log lines that would normally be written
3362 both the the main log and to the reject log were not written to syslog at
3365 PH/14 In the default configuration, change the use of "message" in ACL warn
3366 statements to "add_header".
3368 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3369 not followed by a command (e.g. "seen endif").
3371 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3372 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3375 PH/17 Added extra conditions to the default value of delay_warning_condition
3379 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3380 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3381 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3384 The Auto-Submitted: and various List- headers are standardised, whereas I
3385 don't think Precedence: ever was.
3387 PH/18 Refactored debugging code in route_finduser() to show more information,
3388 in particular, the error code if getpwnam() issues one.
3390 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3391 This is apparently needed in addition to the PH/07 change above to avoid
3392 any possible encoding problems.
3394 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3395 but not after initializing Perl.
3397 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3398 output them only if debugging. By default they are written stderr,
3399 apparently, which is not desirable.
3401 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3404 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3407 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3409 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3410 authenticated or an ident call has been made. Suppress the default
3411 values for $authenticated_id and $authenticated_sender (but permit -oMai
3412 and -oMas) when testing with -bh.
3414 PH/24 Re-jigged the order of the tests in the default configuration so that the
3415 tests for valid domains and recipients precede the DNS black list and CSA
3416 tests, on the grounds that those ones are more expensive.
3418 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3419 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3420 This bug exists in every version of Exim that I still have, right back to
3423 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3424 However, an attempt to turn on case-sensitivity in a regex key by
3425 including (?-i) didn't work because the subject string was already
3426 lowercased, and the effects were non-intuitive. It turns out that a
3427 one-line patch can be used to allow (?-i) to work as expected.
3433 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3434 other effects) broke the use of negated acl sub-conditions.
3436 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3439 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3440 "Deny" causes Exim to reject the incoming connection with a 554 error.
3441 Unfortunately, if there is a major crisis, such as a disk failure,
3442 tcp-wrappers gives "deny", whereas what one would like would be some
3443 kind of temporary error. A kludge has been added to help with this.
3444 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3445 554 error is used if errno is still zero or contains ENOENT (which occurs
3446 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3449 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3451 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3452 errors. Otherwise a message that provokes a temporary error (when other
3453 messages do not) can cause a whole host to time out.
3455 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3456 addresses were routed directly to files or pipes from a redirect router.
3457 File deliveries just didn't batch; pipe deliveries might have suffered
3460 PH/06 A failure to get a lock for a hints database would erroneously always say
3461 "Failed to get write lock", even when it was really a read lock.
3463 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3464 of 0600. This has been changed to use the value of the lockfile_mode
3465 option (which defaults to 0600).
3467 PH/08 Applied small patch from the Sieve maintainer.
3469 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3470 folder from quota calculations, a direct delivery into this folder messed
3471 up the contents of the maildirsize file. This was because the regex was
3472 used only to exclude .Trash (or whatever) when the size of the mailbox
3473 was calculated. There was no check that a delivery was happening into an
3474 excluded directory. This bug has been fixed by ignoring all quota
3475 processing for deliveries into excluded directories.
3477 PH/10 Added the maildirfolder_create_regex option to appendfile.
3483 PH/01 The code for finding all the local interface addresses on a FreeBSD
3484 system running IPv6 was broken. This may well have applied to all BSD
3485 systems, as well as to others that have similar system calls. The broken
3486 code found IPv4 interfaces correctly, but gave incorrect values for the
3487 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3488 that it would not match correctly against @[] and not recognize the IPv6
3491 PH/02 The ipliteral router was not recognizing addresses of the form user@
3492 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3494 PH/03 Added disable_ipv6.
3496 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3497 decoded form, because it is most often used to construct To: headers
3498 lines in autoreplies, and the decoded form may well be syntactically
3499 invalid. However, $reply_address has leading white space removed, and all
3500 newlines turned into spaces so that the autoreply transport does not
3503 PH/05 If group was specified without a user on a router, and no group or user
3504 was specified on a transport, the group from the router was ignored.
3506 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3507 for visible compile-time settings that can be used to change these
3508 numbers, for those that want even more. Backwards compatibility with old
3509 spool files has been maintained. However, going back to a previous Exim
3510 release will lost any variables that are in spool files.
3512 PH/07 Two small changes when running in the test harness: increase delay when
3513 passing a TCP/IP connection to a new process, in case the original
3514 process has to generate a bounce, and remove special handling of
3515 127.0.0.2 (sic), which is no longer necessary.
3517 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3518 be the same on different OS.
3520 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3523 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3524 whether --show-vars was specified or not
3526 JJ/02 exipick: Added support for new ACL variable spool format introduced
3529 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3530 syntactically invalid From: or Reply-to: line, and a filter used this to
3531 generate an autoreply, and therefore failed to obtain an address for the
3532 autoreply, Exim could try to deliver to a non-existent relative file
3533 name, causing unrelated and misleading errors. What now happens is that
3534 it logs this as a hard delivery error, but does not attempt to create a
3537 PH/11 The exinext utility has a -C option for testing purposes, but although
3538 the given file was scanned by exinext itself; it wasn't being passed on
3539 when Exim was called.
3541 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3542 an end-of-file indication when reading a command response.
3544 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3545 compiled. In many other places in Exim, IPv6 addresses are always
3546 recognized, so I have changed this. It also means that IPv4 domain
3547 literals of the form [IPV4:n.n.n.n] are now always recognized.
3549 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3550 used if the router is not running as root, for example, when verifying at
3551 ACL time, or when using -bh. The debugging output from this situation was
3552 non-existent - all you got was a failure to exec. I have made two
3555 (a) Failures to set uid/gid, the current directory, or a process leader
3556 in a subprocess such as that created by queryprogram now generate
3557 suitable debugging output when -d is set.
3559 (b) The queryprogram router detects when it is not running as root,
3560 outputs suitable debugging information if -d is set, and then runs
3561 the subprocess without attempting to change uid/gid.
3563 PH/15 Minor change to Makefile for building test_host (undocumented testing
3566 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3567 additional section of a DNS packet that returns MX or SRV records.
3568 Instead, it always explicitly searches for A/AAAA records. This avoids
3569 major problems that occur when a DNS server includes only records of one
3570 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3571 fixed another bug: if SRV records were looked up and the corresponding
3572 address records were *not* found in the additional section, the port
3573 values from the SRV records were lost.
3575 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3576 using the correct key (the original address) when searching the retry
3577 rules in order to find which one to use for generating the retry hint.
3579 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3580 adding the default one. Similarly, if it contains a Reply-To: header, the
3581 errors_reply_to option, if set, is not used.
3583 PH/19 When calculating a retry time, Exim used to measure the "time since
3584 failure" by looking at the "first failed" field in the retry record. Now
3585 it does not use this if it is later than than the arrival time of the
3586 message. Instead it uses the arrival time. This makes for better
3587 behaviour in cases where some deliveries succeed, thus re-setting the
3588 "first failed" field. An example is a quota failure for a huge message
3589 when small messages continue to be delivered. Without this change, the
3590 "time since failure" will always be short, possible causing more frequent
3591 delivery attempts for the huge message than are intended.
3592 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3594 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3595 $1, $2, $3) because the numerical variables can be reset during some
3596 expansion items (e.g. "match"), thereby losing the authentication data.
3598 PH/21 Make -bV show the size of off_t variables so that the test suite can
3599 decide whether to run tests for quotas > 2G.
3601 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3602 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3603 filecount value is greater than 2G or if a quota value is greater than 2G
3604 on a system where the size of off_t is not greater than 4, a panic error
3607 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3608 never match. The debug and -bh output now contains an explicit error
3609 message indicating a malformed IPv4 address or mask.
3611 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3612 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3613 PH/23 above applies.
3615 PH/25 Do not write to syslog when running in the test harness. The only
3616 occasion when this arises is a failure to open the main or panic logs
3617 (for which there is an explicit test).
3619 PH/26 Added the /no_tell option to "control=freeze".
3621 PH/27 If a host name lookup failed very early in a connection, for example, if
3622 the IP address matched host_lookup and the reverse lookup yielded a name
3623 that did not have a forward lookup, an error message of the form "no IP
3624 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3625 could be logged. Now it outputs the IP address instead of "NULL".
3627 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3628 allows the sender and the authenticated sender to be set when
3629 submitting a message from within Exim. Since child_open_exim() is
3630 documented for local_scan(), the new function should be too.
3632 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3633 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3634 results in an empty string is now treated as unset.
3636 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3638 PH/31 Added sender_verify_fail logging option.
3640 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3641 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3642 tidied the source and removed it altogether.
3644 PH/33 When a queue run was abandoned because the load average was too high, a
3645 log line was always written; now it is written only if the queue_run log
3646 selector is set. In addition, the log line for abandonment now contains
3647 information about the queue run such as the pid. This is always present
3648 in "start" and "stop" lines but was omitted from the "abandon" line.
3650 PH/34 Omit spaces between a header name and the colon in the error message that
3651 is given when verify = headers_syntax fails (if there are lots of them,
3652 the message gets confusing).
3654 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3655 names, as there are now some PTR records that contain slashes. This check
3656 is only to protect against broken name servers that fall over on strange
3657 characters, so the fact that it applies to all lookups doesn't matter.
3659 PH/36 Now that the new test suite is complete, we can remove some of the
3660 special code in Exim that was needed for the old test suite. For example,
3661 sorting DNS records because real resolvers return them in an arbitrary
3662 order. The new test suite's fake resolver always returns records in the
3665 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3666 bounces) except when queue_only is set, to avoid logging races between
3667 the different processes.
3669 PH/38 Panic-die if .include specifies a non-absolute path.
3671 PH/39 A tweak to the "H" retry rule from its user.
3673 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3674 a label. They prevented compilation on older perls.
3676 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3677 a warning to be raised on newish perls.
3679 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3680 on queue. Changes to match documented behaviour of showing count of
3681 messages matching specified criteria.
3683 PH/40 Changed the default ident timeout from 30s to 5s.
3685 PH/41 Added support for the use of login_cap features, on those BSD systems
3686 that have them, for controlling the resources used by pipe deliveries.
3688 PH/42 The content-scanning code uses fopen() to create files in which to put
3689 message data. Previously it was not paying any attention to the mode of
3690 the files. Exim runs with umask(0) because the rest of the code creates
3691 files with open(), and sets the required mode explicitly. Thus, these
3692 files were ending up world-writeable. This was not a big issue, because,
3693 being within the spool directory, they were not world-accessible. I have
3694 created a function called modefopen, which takes an additional mode
3695 argument. It sets umask(777), creates the file, chmods it to the required
3696 mode, then resets the umask. All the relevant calls to fopen() in the
3697 content scanning code have been changed to use this function.
3699 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3700 to 24 hours. This avoids potential overflow problems when processing G
3701 and H retry rules. I suspect nobody ever tinkers with this value.
3703 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3705 PH/45 When the plaintext authenticator is running as a client, the server's
3706 challenges are checked to ensure they are valid base64 strings. By
3707 default, the authentication attempt is cancelled if an invalid string is
3708 received. Setting client_ignore_invalid_base64 true ignores these errors.
3709 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3710 they are received. Thus, the responses can be made to depend on the
3711 challenges. If an invalid string is ignored, an empty string is placed in
3714 PH/46 Messages that are created by the autoreply transport now contains a
3715 References: header, in accordance with RFCs 2822 and 3834.
3717 PH/47 Added authenticated_sender_force to the smtp transport.
3719 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3721 PH/49 Installed latest patch from the Sieve maintainer.
3723 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3724 was also set, the appendfile transport was unnecessarily scanning a
3725 directory of message files (e.g. for maildir delivery) to find the count
3726 of files (along with the size), even though it did not need this
3727 information. It now does the scan only if it needs to find either the
3728 size of the count of files.
3730 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3732 PH/52 Two bugs concerned with error handling when the smtp transport is
3735 (i) Exim was not creating retry information for temporary errors given
3736 for individual recipients after the DATA command when the smtp transport
3737 was used in LMTP mode. This meant that they could be retried too
3738 frequently, and not timed out correctly.
3740 (ii) Exim was setting the flag that allows error details to be returned
3741 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3742 recipients that were returned after the DATA command.
3744 PH/53 This is related to PH/52, but is more general: for any failing address,
3745 when detailed error information was permitted to be returned to the
3746 sender, but the error was temporary, then after the final timeout, only
3747 "retry timeout exceeded" was returned. Now it returns the full error as
3748 well as "retry timeout exceeded".
3750 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3751 do this, and (what is worse) MTAs that accept it.
3753 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3754 will now be deprecated.
3756 PH/56 New os.c-cygwin from the Cygwin maintainer.
3758 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3759 formats (previously only available in exim formats via -bpr, -bpru,
3760 and -bpra. Now also available in native and exiqgrep formats)
3762 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3763 with very large, slow to parse queues
3765 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3767 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3769 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3770 responses to authentication challenges, though it was showing the
3771 challenges; (ii) I've removed the CR characters from the debug output for
3774 PH/58 Allow for the insertion of a newline as well as a space when a string
3775 is turned into more than one encoded-word during RFC 2047 encoding. The
3776 Sieve code now uses this.
3778 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3779 data_4xx, lost_connection, tls_required.
3781 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3782 message was being sent as an SMTP response.
3784 PH/61 Add -l and -k options to exicyclog.
3786 PH/62 When verifying, if an address was redirected to one new address, so that
3787 verification continued, and the new address failed or deferred after
3788 having set something in $address_data, the value of $address_data was not
3789 passed back to the ACL. This was different to the case when no
3790 redirection occurred. The value is now passed back in both cases.
3792 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3793 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3794 use login_cap.h, so on its own it isn't the distinguishing feature. The
3795 new name refers directly to the setclassresources() function.
3797 PH/65 Added configuration files for NetBSD3.
3799 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3801 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3802 is preferred over IPv4.
3804 PH/68 The bounce_return_message and bounce_return_body options were not being
3805 honoured for bounces generated during the reception of non-SMTP messages.
3806 In particular, this applied to messages rejected by the ACL. This bug has
3807 been fixed. However, if bounce_return_message is true and bounce_return_
3808 body is false, the headers that are returned for a non-SMTP message
3809 include only those that have been read before the error was detected.
3810 (In the case of an ACL rejection, they have all been read.)
3812 PH/69 The HTML version of the specification is now built in a directory called
3813 spec_html instead of spec.html, because the latter looks like a path with
3814 a MIME-type, and this confuses some software.
3816 PH/70 Catch two compiler warnings in sieve.c.
3818 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3819 function verify_get_ident() calls ip_connect() to connect a socket, but
3820 if the "connect()" function timed out, ip_connect() used to close the
3821 socket. However, verify_get_ident() also closes the socket later, and in
3822 between Exim writes to the log, which may get opened at this point. When
3823 the socket was closed in ip_connect(), the log could get the same file
3824 descriptor number as the socket. This naturally causes chaos. The fix is
3825 not to close the socket in ip_connect(); the socket should be closed by
3826 the function that creates it. There was only one place in the code where
3827 this was missing, in the iplookup router, which I don't think anybody now
3828 uses, but I've fixed it anyway.
3830 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3831 well as to direct DNS lookups. Otherwise the handling of names in host
3832 lists is inconsistent and therefore confusing.
3838 PH/01 Two changes to the default runtime configuration:
3840 (1) Move the checks for relay_from_hosts and authenticated clients from
3841 after to before the (commented out) DNS black list checks.
3843 (2) Add control=submission to the relay_from_hosts and authenticated
3844 clients checks, on the grounds that messages accepted by these
3845 statements are most likely to be submissions.
3847 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3849 (1) Generate an error if the third argument for the ${prvs expansion is
3852 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3855 (3) Reset the variables that are obtained from the first argument of
3856 ${prvscheck and used in the second argument before leaving the code,
3857 because their memory is reclaimed, so using them afterwards may do
3860 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3861 one (it's much easier than Tom thought :-).
3863 (5) Because of (4), we can now allow for the use of $prvscheck_result
3864 inside the third argument.
3866 PH/03 For some reason, the default setting of PATH when running a command from
3867 a pipe transport was just "/usr/bin". I have changed it to
3870 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3871 anything to be listed in the output from -bV.
3873 PH/05 When a filter generated an autoreply, the entire To: header line was
3874 quoted in the delivery log line, like this:
3876 => >A.N.Other <ano@some.domain> <original@ddress> ...
3878 This has been changed so that it extracts the operative address. There
3879 may be more than one such address. If so, they are comma-separated, like
3882 => >ano@some.domain,ona@other.domain <original@ddress> ...
3884 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3885 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3886 not being looked up in the rDNS to get a host name, Exim was showing the
3887 IP address twice in Received: lines, even though the IP addresses were
3888 identical. For example:
3890 Received: from [1.2.3.4] (helo=[1.2.3.4])
3892 However, if the real host name was known, it was omitting the HELO data
3893 if it matched the actual IP address. This has been tidied up so that it
3894 doesn't show the same IP address twice.
3896 PH/07 When both +timestamp and +memory debugging was on, the value given by
3897 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3898 called by the debug printing, thereby overwriting the timestamp buffer.
3899 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3901 PH/08 When the original message was included in an autoreply transport, it
3902 always said "this is a copy of the message, including all the headers",
3903 even if body_only or headers_only was set. It now gives an appropriate
3906 PH/09 Applied a patch from the Sieve maintainer which:
3908 o fixes some comments
3909 o adds the (disabled) notify extension core
3910 o adds some debug output for the result of if/elsif tests
3911 o points to the current vacation draft in the documentation
3912 and documents the missing references header update
3916 o fixes a bug in processing the envelope test (when testing
3917 multiple envelope elements, the last element determined the
3920 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3921 Electronic Mail") by including:
3923 Auto-submitted: auto-generated
3925 in the messages that it generates (bounce messages and others, such as
3926 warnings). In the case of bounce messages for non-SMTP messages, there was
3927 also a typo: it was using "Auto_submitted" (underscore instead of
3928 hyphen). Since every message generated by Exim is necessarily in response
3929 to another message, thes have all been changed to:
3931 Auto-Submitted: auto-replied
3933 in accordance with these statements in the RFC:
3935 The auto-replied keyword:
3937 - SHOULD be used on messages sent in direct response to another
3938 message by an automatic process,
3940 - MUST NOT be used on manually-generated messages,
3942 - MAY be used on Delivery Status Notifications (DSNs) and Message
3943 Disposition Notifications (MDNs),
3945 - MUST NOT be used on messages generated by automatic or periodic
3946 processes, except for messages which are automatic responses to
3949 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3950 to the default Received: header definition.
3952 PH/12 Added log selector acl_warn_skipped (default on).
3954 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3955 variables because (a) they are in the wrong storage pool and (b) even if
3956 they were copied, it wouldn't work properly because of the caching.
3958 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3959 checking when decoding. Apparently there are clients that generate
3960 overlong encoded strings. Why am I not surprised?
3962 PH/15 If the first argument of "${if match_address" was not empty, but did not
3963 contain an "@" character, Exim crashed. Now it writes a panic log message
3964 and treats the condition as false.
3966 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3968 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3969 extension "envelope-auth". The code is finished and in agreement with
3970 other implementations, but there is no documentation so far and in fact,
3971 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3972 not changing the active code.
3974 Print executed "if" and "elsif" statements when debugging is used. This
3975 helps a great deal to understand what a filter does.
3977 Document more things not specified clearly in RFC3028. I had all this
3978 sorted out, when out of a sudden new issues came to my mind. Oops."
3980 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3983 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3984 canonical form (as documented). However, after a host name lookup from
3985 the IP address, check_host() was doing a simple string comparison with
3986 addresses acquired from the DNS when checking that the found name did
3987 have the original IP as one of its addresses. Since any found IPv6
3988 addresses are likely to be in abbreviated form, the comparison could
3989 fail. Luckily, there already exists a function for doing the comparison
3990 by converting both addresses to binary, so now that is used instead of
3991 the text comparison.
3993 PH/20 There was another similar case to PH/19, when a complete host name was
3994 given in a host list; looking up its IP address could give an abbreviated
3995 form, whereas the current host's name might or might not be abbreviated.
3996 The same fix has been applied.
4002 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4003 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4006 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4007 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4009 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4011 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4012 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4013 Also, for "Auto-Submitted", treat anything other than "no" as
4014 non-personal, in accordance with RFC 3834. (Previously it treated
4015 anything starting "auto-" as non-personal.)
4017 TF/01 The control=submission/name=... option had a problem with syntax
4018 errors if the name included a slash character. The /name= option
4019 now slurps the rest of the string, so it can include any characters
4020 but it must come last in the list of options (after /sender_retain
4023 PH/05 Some modifications to the interface to the fake nameserver for the new
4031 TK/01 Added the "success_on_redirect" address verification option. See
4032 NewStuff for rationale and an example.
4034 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4036 PH/02 Patch to exigrep to allow it to work on syslog lines.
4038 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4039 fread() to skip over the body file's header line, because in Cygwin the
4040 header line is locked and is inaccessible.
4042 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4043 co-exist for some time) to make it clear that it is the Exim ID that is
4044 referenced, not the Message-ID: header line.
4046 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4047 string_format() function, because snprintf() does not exist on all
4050 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4051 ${stat: expansion item.
4053 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4054 protocol synchronization error", to keep the pedants happy.
4056 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4057 well as for IRIX systems, when gcc is being used. See the host.c source
4060 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4062 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4065 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4066 transport and to the smtp transport in LMTP mode.
4068 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4070 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4071 run a filter in a subprocess. This could lead to confusion in subsequent
4072 lookups in the parent process. There should also be a search_tidyup() at
4073 the end of the subprocess.
4075 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4076 only if the host matched helo_try_verify_hosts, which caused the
4077 verification to occur when the EHLO/HELO command was issued. The ACL just
4078 tested the remembered result. Now, if a previous verification attempt has
4079 not happened, "verify = helo" does it there and then.
4081 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4083 TK/03 Fix log output including CR from clamd.
4085 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4086 exist provoked a memory error which could cause a segfault.
4088 PH/15 Installed PCRE 6.2
4090 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4092 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4093 of the problem. Specifically, suggested +O2 rather than +O1 for the
4096 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4098 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4099 record, the port was not correctly logged when the outgoing_port log
4100 selector was set (it logged the transort's default port).
4102 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4103 fallback_hosts, and "hosts" in the smtp transport.
4105 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4106 host errors such as "Connection refused".
4108 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4109 authentication with radiusclient 0.4.9:
4111 - Error returned from rc_read_config was caught wrongly
4112 - Username/password not passed on to radius server due to wrong length.
4114 The presumption is that some radiusclient API changes for 4.51/PH/17
4115 were not taken care of correctly. The code is still untested by me (my
4116 Linux distribution still has 0.3.2 of radiusclient), but it was
4117 contributed by a Radius user.
4119 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4120 expanding the "port" option of the smtp transport.
4122 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4123 while reading a MIME header. Thanks to Tom Hughes for a patch.
4125 PH/24 Include config.h inside local_scan.h so that configuration settings are
4128 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4129 that in an ACL for RCPT (for example), you can examine exactly what was
4132 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4133 commands, but it was not correctly comparing the address with the actual
4134 client host address. Thus, it would show the EHLO address in Received:
4135 header lines when this was not necessary.
4137 PH/27 Added the % operator to ${eval:}.
4139 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4140 it should be ignoring failures (because with -C, for example, it has lost
4141 privilege). It wasn't ignoring creation failures other than "already
4144 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4147 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4148 because an input line was too long, either on its own, or by virtue of
4149 too many continuations, the temporary file was not being removed, and the
4150 return code was incorrect.
4152 PH/31 Missing "BOOL" in function definition in filtertest.c.
4154 PH/32 Applied Sieve patches from the maintainer.
4156 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4158 PH/33 Added "verify = not_blind".
4160 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4161 Local/Makefile (with some defaults set). These are used in built scripts
4162 such as exicyclog, but they have never been used in the exim_install
4163 script (though there are many overriding facilities there). I have
4164 arranged that the exim_install script now takes note of these two
4167 PH/35 Installed configuration files for Dragonfly.
4169 PH/36 When a locally submitted message by a trusted user did not contain a
4170 From: header, and the sender address was obtained from -f or from an SMTP
4171 MAIL command, and the trusted user did not use -F to supply a sender
4172 name, $originator_name was incorrectly used when constructing a From:
4173 header. Furthermore, $originator_name was used for submission mode
4174 messages from external hosts without From: headers in a similar way,
4175 which is clearly wrong.
4177 PH/37 Added control=suppress_local_fixups.
4179 PH/38 When log_selector = +received_sender was set, and the addition of the
4180 sender made the log line's construction buffer exactly full, or one byte
4181 less than full, an overflow happened when the terminating "\n" was
4184 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4185 when the result of a list match is failure because a DNS lookup failed.
4187 PH/40 RM_COMMAND is now used in the building process.
4189 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4190 the "build-* directories that it finds.
4192 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4193 address in a domain literal was a prefix of an interface address.
4195 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4196 when verifying a sender address, unless rewrite_headers is false.
4198 PH/44 Wrote a long comment about why errors_to addresses are verified as
4199 recipients, not senders.
4201 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4202 the ratelimit ACL was added.
4204 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4206 PH/46 Added extra information about PostgreSQL errors to the error string.
4208 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4209 avoiding the need to install special zones in a real server. This is
4210 backwards compatible; if it can't find the fake resolver, it drops back.
4211 Thus, both old and new test suites can be run.
4213 TF/02 Added util/ratelimit.pl
4215 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4216 clock is set back in time.
4218 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4219 Candler <B.Candler@pobox.com>.
4221 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4222 OK for addresses that are the result of redirections.
4224 PH/48 A number of further additions for the benefit of the new test suite,
4225 including a fake gethostbyname() that interfaces to the fake DNS resolver
4228 TF/06 The fix for widen_domains has also been applied to qualify_single and
4229 search_parents which are the other dnslookup options that can cause
4232 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4235 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4237 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4238 service" if it encountered an MX record with an empty target hostname.
4239 The message is now "an MX or SRV record indicated no SMTP service".
4241 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4242 if the DNS of the sending site is misconfigured. This is quite a
4243 common situation. This change restores the behaviour of treating a
4244 helo verification defer as a failure.
4246 PH/51 If self=fail was set on a router, the bounce message did not include the
4247 actual error message.
4253 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4255 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4256 command itself ended in error, the underlying message about the transport
4257 filter timeout was being overwritten with the pipe command error. Now the
4258 underlying error message should be appended to the second error message.
4260 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4262 PH/02 Reduce the amount of output that "make" produces by default. Full output
4263 can still be requested.
4265 PH/03 The warning log line about a condition test deferring for a "warn" verb
4266 was being output only once per connection, rather than after each
4267 occurrence (because it was using the same function as for successful
4268 "warn" verbs). This seems wrong, so I have changed it.
4270 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4271 it should not have, which might have caused a crash in the right
4272 circumstances, but probably never did.
4274 PH/04 Installed a modified version of Tony Finch's patch to make submission
4275 mode fix the return path as well as the Sender: header line, and to
4276 add a /name= option so that you can make the user's friendly name appear
4279 TF/03 Added the control = fakedefer ACL modifier.
4281 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4282 Mark Lowes for thorough testing.
4284 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4286 TK/03 Merged latest SRS patch from Miles Wilton.
4288 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4289 with the definition in sysexits.h (which is #included earlier).
4290 Fortunately, Exim does not actually use EX_OK. The code used to try to
4291 preserve the sysexits.h value, by assuming that macro definitions were
4292 scanned for macro replacements. I have been disabused of this notion,
4293 so now the code just undefines EX_OK before #including unistd.h.
4295 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4296 in the smtp transport. When a block could not be written in a single
4297 write() function, the timeout was being re-applied to each part-write.
4298 This seems wrong - if the receiver was accepting one byte at a time it
4299 would take for ever. The timeout is now adjusted when this happens. It
4300 doesn't have to be particularly precise.
4302 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4303 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4305 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4306 without a domain if the check to <postmaster@domain> fails.
4308 SC/01 Eximstats: added -xls and the ability to specify output files
4309 (patch written by Frank Heydlauf).
4311 SC/02 Eximstats: use FileHandles for outputting results.
4313 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4315 SC/04 Eximstats: fixed display of large numbers with -nvr option
4317 SC/05 Eximstats: fixed merging of reports with empty tables.
4319 SC/06 Eximstats: added the -include_original_destination flag
4321 SC/07 Eximstats: removed tabs and trailing whitespace.
4323 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4325 TK/06 MBOX spool code: Add real "From " MBOX separator line
4326 so the .eml file is really in mbox format (even though
4327 most programs do not really care). Patch from Alex Miller.
4329 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4330 The latter is generated from $received_to and is only set if the
4331 message has one envelope recipient. SA can use these headers,
4332 obviously out-of-the-box. Patch from Alex Miller.
4334 PH/08 The ${def test on a variable was returning false if the variable's
4335 value was "0", contrary to what the specification has always said!
4336 The result should be true unless the variable is empty.
4338 PH/09 The syntax error of a character other than { following "${if
4339 def:variable_name" (after optional whitespace) was not being diagnosed.
4340 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4341 accidental colon was present, for example, could give incorrect results.
4343 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4344 result is used (not including appendfile, where other changes are about
4347 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4348 This involved changing a lot of size variables from int to off_t. It
4349 should work with maildirs and everything.
4351 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4352 spamd dying while we are connected to it.
4354 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4357 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4358 function for BDB 4.3.
4360 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4362 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4363 the sources for PCRE have been re-arranged and more files are now
4366 PH/15 The code I had for printing potentially long long variables in PH/11
4367 above was not the best (it lost precision). The length of off_t variables
4368 is now inspected at build time, and an appropriate printing format (%ld
4369 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4370 to be "long long int" or "long int". This is needed for the internal
4371 formatting function string_vformat().
4373 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4374 the configuration file to be ":syslog", then the script "guesses" where
4375 the logs files are, rather than using the compiled in default. In our
4376 case the guess is not the same as the compiled default, so the script
4377 suddenly stopped working when I started to use syslog. The patch checks
4378 to see if log_file_path is "". If so, it attempts to read it from exim
4379 with no configuration file to get the compiled in version, before it
4380 falls back to the previous guessing code."
4382 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4383 implementing BATV in an Exim configuration. See NewStuff for the gory
4386 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4387 Makefile that are specific to HP-UX.
4389 PH/18 If the "use_postmaster" option was set for a recipient callout together
4390 with the "random" option, the postmaster address was used as the MAIL
4391 FROM address for the random test, but not for the subsequent recipient
4392 test. It is now used for both.
4394 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4395 patch removes a few documentation additions to RFC 3028, because the
4396 latest draft now contains them. It adds the new en;ascii-case comparator
4397 and a new error check for 8bit text in MIME parts. Comparator and
4398 require names are now matched exactly. I enabled the subaddress
4399 extension, but it is not well tested yet (read: it works for me)."
4401 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4402 rework some of the code of TK/09 above to avoid the hardwired use of
4403 "%lld" and "long long". Replaced the call to snprintf() with a call to
4406 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4407 records point to non-existent hosts", "retry timeout exceeded", and
4408 "retry time not reached for any host after a long failure period".
4410 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4411 experimental DomainKeys support:
4413 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4414 (2) On an error such as an illegally used "control", the wrong name for
4415 the control was given.
4417 These problems did NOT occur unless DomainKeys support was compiled.
4419 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4421 PH/24 Added ${if match_ip condition.
4423 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4424 databases so that it will be absolutely obvious if a crash occurs in the
4425 DB library. This is a regular occurrence (often caused by mis-matched
4428 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4429 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4430 picked up on a user's system that detects such things. There doesn't seem
4431 to be a gcc warning option for this - only an attribute that has to be
4432 put on the function's prototype. It seems that in Fedora Core 4 they have
4433 set this on a number of new functions. No doubt there will be more in due
4436 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4437 specify a transport. However, if an address that was verified by such a
4438 router was the subject of a callout, Exim crashed because it tried to
4439 read the rcpt_include_affixes from the non-existent transport. Now it
4440 just assumes that the setting of that option is false. This bug was
4441 introduced by 4.51/PH/31.
4443 PH/28 Changed -d+all to exclude +memory, because that information is very
4444 rarely of interest, but it makes the output a lot bigger. People tend to
4445 do -d+all out of habit.
4447 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4448 code in os-type was giving problems when libc.so lives in lib64, like on
4451 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4452 aren't the modern standard, and it seems that some systems' include files
4453 don't always have them. Exim was already checking for some of the newer
4454 ones like T_AAAA, and defining it itself. I've added checks for all the
4455 record types that Exim uses.
4457 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4458 not automatically generating a new one, as it is supposed to. This
4459 prevented TLS from working. If the file did exist, but contained invalid
4460 data, a new version was generated, as expected. It was only the case of a
4461 non-existent file that was broken.
4463 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4464 with a change in libdomainkeys > 0.64.
4466 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4467 from DNS. If the selector record carries the flag, it now has
4468 precedence over the domain-wide flag.
4470 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4472 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4473 the use of a port name that isn't defined in /etc/services) occurred, the
4474 message was deferred as in a normal delivery, and thus remained on the
4475 spool, instead of being failed because of the mua_wrapper setting. This
4476 is now fixed, and I tidied up some of the mua_wrapper messages at the
4479 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4480 the messages in a hash of arrays rather than using individual hashes.
4481 This is a bit cleaner and results in dramatic memory savings, albeit
4482 at a slight CPU cost.
4484 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4485 as requested by Marc Sherman.
4487 SC/10 Eximstats: added histograms for user specified patterns as requested
4490 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4492 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4493 fopen() in the content-scanning modules that did not already have it.
4499 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4500 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4502 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4504 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4506 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4507 to test Sieve filters that use "vacation".
4509 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4510 that changes the way the GnuTLS parameters are stored in the cache file.
4511 The new format can be generated externally. For backward compatibility,
4512 if the data in the cache doesn't make sense, Exim assumes it has read an
4513 old-format file, and it generates new data and writes a new file. This
4514 means that you can't go back to an older release without removing the
4517 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4518 work if there are any delivery delays because "one_time" forces the
4519 parent to be marked "delivered", so its unseen clone is never tried
4520 again. For this reason, Exim now forbids the simultaneous setting of
4523 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4524 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4525 entry if you want to know the details. The fix, however, neglected to
4526 consider the case when local delivery batching is involved. The test for
4527 "previously delivered" was not happening when checking to see if an
4528 address could be batched with a previous (undelivered) one; under
4529 certain circumstances this could lead to multiple deliveries to the same
4532 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4533 in its include files, and this causes problems building Exim.
4535 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4536 header_syntax) but Exim was just ignoring anything given after a slash.
4537 In particular, this caused confusion with an attempt to use "verify =
4538 reverse_host_lookup/defer_ok". An error is now given when options are
4539 supplied for verify items that do not have them. (Maybe reverse_host_
4540 lookup should have a defer_ok option, but that's a different point.)
4542 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4543 defined by RFC 821) to 2048, because there were problems with some AUTH
4544 commands, and RFC 1869 says the size should be increased for extended
4545 SMTP commands that take arguments.
4547 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4550 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4551 "unknown" error; now it says that the functionality isn't in the binary.
4553 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4554 an address' error message when a string expansion fails (syntax or
4555 whatever). Otherwise the password may appear in the log. Following change
4556 PH/42 below, there is no longer a chance of it appearing in a bounce
4559 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4561 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4562 of its addresses was ever tried. (Bugzilla bug #2.)
4564 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4565 the result incorrectly in the debug output. (It correctly added a newline
4566 to what was transported.)
4568 TF/01 Added $received_time.
4570 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4571 commented out examples of how to interface to a virus scanner and to
4572 SpamAssassin. Also added commented examples of av_scanner and
4573 spamd_address settings.
4575 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4576 and controls are allowed in which ACLs. There were a couple of minor
4577 errors. Some of the entries in the conditions table (which is a table of
4578 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4579 negation of where the condition IS allowed.
4581 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4583 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4584 header file does not have a version number, so I've had to invent a new
4585 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4586 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4587 radiusclient), but it was contributed by a Radius user.
4589 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4590 files or directories, for OpenSSL.
4592 PH/19 When an Exim process that is running as root has to create an Exim log
4593 file, it does so in a subprocess that runs as exim:exim so as to get the
4594 ownership right at creation (otherwise, other Exim processes might see
4595 the file with the wrong ownership). There was no test for failure of this
4596 fork() call, which would lead to the process getting stuck as it waited
4597 for a non-existent subprocess. Forks do occasionally fail when resources
4598 run out. I reviewed all the other calls to fork(); they all seem to check
4601 PH/20 When checking for unexpected SMTP input at connect time (before writing
4602 the banner), Exim was not dealing correctly with a non-positive return
4603 from the read() function. If the client had disconnected by this time,
4604 the result was a log entry for a synchronization error with an empty
4605 string after "input=" when read() returned zero. If read() returned -1
4606 (an event I could not check), uninitialized data bytes were printed.
4607 There were reports of junk text (parts of files, etc) appearing after
4610 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4612 PH/22 Added support for macro redefinition, and (re)definition in between
4613 driver and ACL definitions.
4615 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4616 forgetting to use the resulting value; it was using the unexpanded value.
4618 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4619 hadn't been configured. The fix is from Juergen Kreileder, who
4620 understands it better than I do:
4622 "Here's what I see happening with three configured cyrus_sasl
4623 authenticators configured (plain, login, cram-md5):
4625 On startup auth_cyrus_sasl_init() gets called for each of these.
4626 This means three calls to sasl_listmech() without a specified mech_list.
4627 => SASL tests which mechs of all available mechs actually work
4628 => three warnings about OTP not working
4629 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4631 With the patch, sasl_listmech() also gets called three times. But now
4632 SASL's mech_list option is set to the server_mech specified in the the
4633 authenticator. Or in other words, the answer from sasl_listmech()
4634 gets limited to just the mech you're testing for (which is different
4636 => the return list contains just 'plain' or 'login', 'cram-md5' or
4637 nothing depending on the value of ob->server_mech.
4639 I've just tested the patch: Authentication still works fine,
4640 unavailable mechs specified in the exim configuration are still
4641 caught, and the auth.log warnings about OTP are gone."
4643 PH/25 When debugging is enabled, the contents of the command line are added
4644 to the debugging output, even when log_selector=+arguments is not
4647 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4648 answer is "GNU", and only if the return is "GNU/something" is the answer
4651 PH/27 $acl_verify_message is now set immediately after the failure of a
4652 verification in an ACL, and so is available in subsequent modifiers. In
4653 particular, the message can be preserved by coding like this:
4655 warn !verify = sender
4656 set acl_m0 = $acl_verify_message
4658 Previously, $acl_verify_message was set only while expanding "message"
4659 and "log_message" when a very denied access.
4661 PH/28 Modified OS/os.c-Linux with
4663 -#ifndef OS_LOAD_AVERAGE
4664 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4666 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4667 nomenclature these days.)
4669 PH/29 Installed patch from the Sieve maintainer that adds the options
4670 sieve_useraddress and sieve_subaddress to the redirect router.
4672 PH/30 In these circumstances:
4673 . Two addresses routed to the same list of hosts;
4674 . First host does not offer TLS;
4675 . First host accepts first address;
4676 . First host gives temporary error to second address;
4677 . Second host offers TLS and a TLS session is established;
4678 . Second host accepts second address.
4679 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4680 and peerdn, if requested) that were in fact used only for the second
4683 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4684 attention to any local part prefix or suffix that was matched by the
4685 router that accepted the address. It now behaves in the same way as it
4686 does for delivery: the affixes are removed from the local part unless
4687 rcpt_include_affixes is set on the transport.
4689 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4690 timeout during the DATA phase of an incoming message.
4692 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4693 applied a patch sanctioned by the Sieve maintainer.
4695 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4696 the uid or gid is negative. A case of a negative gid caused this to be
4697 noticed. The fix allows for either to be negative.
4699 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4700 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4703 PH/36 The change PH/12 above was broken. Fixed it.
4705 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4706 the grounds that routing the same address twice would always produce the
4707 same answer. This might have been true once, but it is certainly no
4708 longer true now. Routing a child address may depend on the previous
4709 routing that produced that child. Some complicated redirection strategies
4710 went wrong when messages had multiple recipients, and made Exim's
4711 behaviour dependent on the order in which the addresses were given.
4713 I have moved the duplicate checking until after the routing is complete.
4714 Exim scans the addresses that are assigned to local and remote
4715 transports, and removes any duplicates. This means that more work will be
4716 done, as duplicates will always all be routed, but duplicates are
4717 presumably rare, so I don't expect this is of any significance.
4719 For deliveries to pipes, files, and autoreplies, the duplicate checking
4720 still happens during the routing process, since they are not going to be
4723 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4724 It corrects a timeout issue with spamd. This is Ian's comment: "The
4725 background is that sometimes spamd either never reads data from a
4726 connection it has accepted, or it never writes response data. The exiscan
4727 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4728 blindly assumes that writes won't block so it may never time out."
4730 PH/39 Allow G after quota size as well as K and M.
4732 PH/40 The value set for $authenticated_id in an authenticator may not contain
4733 binary zeroes or newlines because the value is written to log lines and
4734 to spool files. There was no check on this. Now the value is run through
4735 the string_printing() function so that such characters are converted to
4736 printable escape sequences.
4738 PH/41 $message_linecount is a new variable that contains the total number of
4739 lines in the message. Compare $body_linecount, which is the count for the
4742 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4743 bounce and delay warning messages, except in certain special cases, which
4746 (a) An SMTP error message from a remote host;
4747 (b) A message specified in a :fail: redirection;
4748 (c) A message specified in a "fail" command in a system filter;
4749 (d) A message specified in a FAIL return from the queryprogram router;
4750 (e) A message specified by the cannot_route_message router option.
4752 In these cases only, Exim does include the error details in bounce and
4753 warning messages. There are also a few cases where bland messages such
4754 as "unrouteable address" or "local delivery error" are given.
4756 PH/43 $value is now also set for the "else" part of a ${run expansion.
4758 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4759 being worked on, but at least Exim now implements the latest version to
4762 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4763 process to complete was treated as a delivery failure, a timeout while
4764 writing the message to the pipe was logged, but erroneously treated as a
4765 successful delivery. Such timeouts include transport filter timeouts. For
4766 consistency with the overall process timeout, these timeouts are now
4767 treated as errors, giving rise to delivery failures by default. However,
4768 there is now a new Boolean option for the pipe transport called
4769 timeout_defer, which, if set TRUE, converts the failures into defers for
4770 both kinds of timeout. A transport filter timeout is now identified in
4773 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4774 systems where "make" and "gmake" are different, calling "gmake" at top
4775 level broke things. I've arranged for the value of $(MAKE) to be passed
4776 from the Makefile to this script so that it can call the same version of
4780 A note about Exim versions 4.44 and 4.50
4781 ----------------------------------------
4783 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4784 changes of various kinds. As a consequence, a big documentation update was
4785 needed. This delayed the release for rather longer than seemed good, especially
4786 in the light of a couple of (minor) security issues. Therefore, the changes
4787 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4788 release. So 4.44 and 4.50 are in effect two different branches that both start
4791 I have left the 4.50 change log unchanged; it contains all the changes since
4792 4.43. The change log for 4.44 is below; many of its items are identical to
4793 those for 4.50. This seems to be the most sensible way to preserve the
4794 historical information.
4800 1. Minor wording change to the doc/README.SIEVE file.
4802 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4803 computation of the current number of files was incorrect.
4805 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4806 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4809 4. Give more explanation in the error message when the command for a transport
4810 filter fails to execute.
4812 5. There are several places where Exim runs a non-Exim command in a
4813 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4814 was being done only for the command run by the queryprogram router. It is
4815 now done for all such subprocesses. The other cases are: ${run, transport
4816 filters, and the commands run by the lmtp and pipe transports.
4818 6. Added CONFIGURE_GROUP build-time option.
4820 7. Some older OS have a limit of 256 on the maximum number of file
4821 descriptors. Exim was using setrlimit() to set 1000 as a large value
4822 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4823 systems. I've change it so that if it can't get 1000, it tries for 256.
4825 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4826 was an oversight, and furthermore, ever since the addition of extra
4827 controls (e.g. 4.43/32), the checks on when to allow different forms of
4828 "control" were broken. There should now be diagnostics for all cases when a
4829 control that does not make sense is encountered.
4831 9. Added the /retain_sender option to "control=submission".
4833 10. $recipients is now available in the predata ACL (oversight).
4835 11. Tidy the search cache before the fork to do a delivery from a message
4836 received from the command line. Otherwise the child will trigger a lookup
4837 failure and thereby defer the delivery if it tries to use (for example) a
4838 cached ldap connection that the parent has called unbind on.
4840 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4841 of $address_data from the recipient verification was clobbered by the
4842 sender verification.
4844 13. The value of address_data from a sender verification is now available in
4845 $sender_address_data in subsequent conditions in the ACL statement.
4847 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4849 15. Added a new option "connect=<time>" to callout options, to set a different
4852 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4853 was its contents. (It was OK if the option was not defined at all.)
4855 17. A "Completed" log line is now written for messages that are removed from
4856 the spool by the -Mrm option.
4858 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4859 information about exactly what failed.
4861 19. Added -dd to debug only the daemon process.
4863 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4864 handles timeouts, both on the server side and network timeouts. Renamed the
4865 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4867 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4868 It is now set to "smtps".
4870 22. $host_address is now set to the target address during the checking of
4871 ignore_target_hosts.
4873 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4874 being passed; this would have caused $sender_host_name to have been used if
4875 matching the list had actually called for a host name (not very likely,
4876 since this list is usually IP addresses). A host name is now passed as
4879 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4880 code that specifies a non-restarting handler (typically sigaction() in
4881 modern systems) in an attempt to fix a rare and obscure crash bug.
4883 25. Narrowed the window for a race in the daemon that could cause it to ignore
4884 SIGCHLD signals. This is not a major problem, because they are used only to
4885 wake it up if nothing else does.
4887 26. A malformed maildirsize file could cause Exim to calculate negative values
4888 for the mailbox size or file count. Odd effects could occur as a result.
4889 The maildirsize information is now recalculated if the size or filecount
4892 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4893 support for a long time. Removed HAVE_SYS_VFS_H.
4895 28. Installed the latest version of exipick from John Jetmore.
4897 29. In an address list, if the pattern was not a regular expression, an empty
4898 subject address (from a bounce message) matched only if the pattern was an
4899 empty string. Non-empty patterns were not even tested. This was the wrong
4900 because it is perfectly reasonable to use an empty address as part of a
4901 database query. An empty address is now tested by patterns that are
4902 lookups. However, all the other forms of pattern expect the subject to
4903 contain a local part and a domain, and therefore, for them, an empty
4904 address still always fails if the pattern is not itself empty.
4906 30. Exim went into a mad DNS loop when attempting to do a callout where the
4907 host was specified on an smtp transport, and looking it up yielded more
4908 than one IP address.
4910 31. Re-factored the code for checking spool and log partition space into a
4911 function that finds that data and another that does the check. The former
4912 is then used to implement four new variables: $spool_space, $log_space,
4913 $spool_inodes, and $log_inodes.
4915 32. The RFC2047 encoding function was originally intended for short strings
4916 such as real names; it was not keeping to the 75-character limit for
4917 encoded words that the RFC imposes. It now respects the limit, and
4918 generates multiple encoded words if necessary. To be on the safe side, I
4919 have increased the buffer size for the ${rfc2047: expansion operator from
4922 33. It is now permitted to omit both strings after an "if" condition; if the
4923 condition is true, the result is "true". As before, when the second string
4924 is omitted, a false condition yields an empty string. This makes it less
4925 cumbersome to write custom ACL and router conditions.
4927 34. Failure to deliver a bounce message always caused it to be frozen, even if
4928 there was an errors_to setting on the router. The errors_to setting is now
4931 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4932 canonical form (fully expanded) before being placed in
4933 $sender_host_address.
4935 36. The table in the code that translates DNS record types into text (T_A to
4936 "A" for instance) was missing entries for NS and CNAME. It is just possible
4937 that this could have caused confusion if both these types were looked up
4938 for the same domain, because the text type is used as part of Exim's
4939 per-process caching. But the chance of anyone hitting this buglet seems
4942 37. The dnsdb lookup has been extended in a number of ways.
4944 (1) There is a new type, "zns", which walks up the domain tree until it
4945 finds some nameserver records. It should be used with care.
4947 (2) There is a new type, "mxh", which is like "mx" except that it returns
4948 just the host names, not the priorities.
4950 (3) It is now possible to give a list of domains (or IP addresses) to be
4951 looked up. The behaviour when one of the lookups defers can be
4952 controlled by a keyword.
4954 (4) It is now possible to specify the separator character for use when
4955 multiple records are returned.
4957 38. The dnslists ACL condition has been extended: it is now possible to supply
4958 a list of IP addresses and/or domains to be looked up in a particular DNS
4961 39. Added log_selector=+queue_time_overall.
4963 40. When running the queue in the test harness, wait just a tad after forking a
4964 delivery process, to get repeatability of debugging output.
4966 41. Include certificate and key file names in error message when GnuTLS fails
4967 to set them up, because the GnuTLS error message doesn't include the name
4968 of the failing file when there is a problem reading it.
4970 42. Allow both -bf and -bF in the same test run.
4972 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4974 44. The "Exiscan patch" is now merged into the mainline Exim source.
4976 45. Sometimes the final signoff response after QUIT could fail to get
4977 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4978 < 0 before doing a fflush(). This bug looks as though it goes back to the
4979 introduction of TLS in release 3.20, but "sometimes" must have been rare
4980 because the tests only now provoked it.
4982 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4983 (this can affect the format of dates).
4985 47. exim_tidydb, when checking for the continued existence of a message for
4986 which it has found a message-specific retry record, was not finding
4987 messages that were in split spool directories. Consequently, it was
4988 deleting retry records that should have stayed in existence.
4990 48. Steve fixed some bugs in eximstats.
4992 49. The SPA authentication driver was not abandoning authentication and moving
4993 on to the next authenticator when an expansion was forced to fail,
4994 contradicting the general specification for all authenticators. Instead it
4995 was generating a temporary error. It now behaves as specified.
4997 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4998 (the order specifies the preference for clients). The order is now AES256,
4999 AES128, 3DES, ARCFOUR128.
5001 51. Small patch to Sieve code - explicitly set From: when generating an
5004 52. Exim crashed if a remote delivery caused a very long error message to be
5005 recorded - for instance if somebody sent an entire SpamAssassin report back
5006 as a large number of 550 error lines. This bug was coincidentally fixed by
5007 increasing the size of one of Exim's internal buffers (big_buffer) that
5008 happened as part of the Exiscan merge. However, to be on the safe side, I
5009 have made the code more robust (and fixed the comments that describe what
5012 53. Now that there can be additional text after "Completed" in log lines (if
5013 the queue_time_overall log selector is set), a one-byte patch to exigrep
5014 was needed to allow it to recognize "Completed" as not the last thing in
5017 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5018 patch that reportedly fixes this has been added. I am not expert enough to
5019 create a test for it. This is what the patch creator wrote:
5021 "I found a little strange behaviour of ldap code when working with
5022 Windows 2003 AD Domain, where users was placed in more than one
5023 Organization Units. When I tried to give exim partial DN, the exit code
5024 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5025 But simultaneously result of request was absolutely normal ldap result,
5026 so I produce this patch..."
5028 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5029 REFERENCE, so I have modified the code to exclude the patch when that macro
5032 55. Some experimental protocols are using DNS PTR records for new purposes. The
5033 keys for these records are domain names, not reversed IP addresses. The
5034 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5035 leaves it alone. Component reversal etc. now happens only for IP addresses.
5038 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5040 57. Double the size of the debug message buffer (to 2048) so that more of very
5041 long debug lines gets shown.
5043 58. The exicyclog utility now does better if the number of log files to keep
5044 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5046 59. Two changes related to the smtp_active_hostname option:
5048 (1) $smtp_active_hostname is now available as a variable.
5049 (2) The default for smtp_banner uses $smtp_active_hostname instead
5050 of $primary_hostname.
5052 60. The host_aton() function is supposed to be passed a string that is known
5053 to be a valid IP address. However, in the case of IPv6 addresses, it was
5054 not checking this. This is a hostage to fortune. Exim now panics and dies
5055 if the condition is not met. A case was found where this could be provoked
5056 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5057 components; fortuitously, this particular loophole had already been fixed
5058 by change 4.50/55 above.
5060 If there are any other similar loopholes, the new check in host_aton()
5061 itself should stop them being exploited. The report I received stated that
5062 data on the command line could provoke the exploit when Exim was running as
5063 exim, but did not say which command line option was involved. All I could
5064 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5065 running as the user.
5068 61. There was a buffer overflow vulnerability in the SPA authentication code
5069 (which came originally from the Samba project). I have added a test to the
5070 spa_base64_to_bits() function which I hope fixes it.
5073 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5074 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5076 63. The daemon start-up calls getloadavg() while still root for those OS that
5077 need the first call to be done as root, but it missed one case: when
5078 deliver_queue_load_max is set with deliver_drop_privilege. This is
5079 necessary for the benefit of the queue runner, because there is no re-exec
5080 when deliver_drop_privilege is set.
5082 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5083 This has been fixed.
5085 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5086 was in use, was not putting the data itself into the right store pool;
5087 consequently, it could be overwritten for a subsequent message in the same
5088 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5091 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5093 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5094 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5095 the function were treating the return as a boolean value, which happened to
5096 work because 0=false and not-0=true, but is not correct code.
5098 68. The host_aton() function was not handling scoped IPv6 addresses (those
5099 with, for example, "%eth0" on the end) correctly.
5101 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5102 negated items (that is, ~something) in unsigned ints. Some compilers
5103 apparently mutter when there is no cast.
5105 70. If an address verification called from an ACL failed, and did not produce a
5106 user-specific message (i.e. there was only a "system" message), nothing was
5107 put in $acl_verify_message. In this situation, it now puts the system
5110 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5111 session; change 4.31/43 added the unwanted input to the log line - except
5112 that it did not do this in the start of session case. It now does.
5114 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5115 This is wrong and can cause the other end to generate a synchronization
5116 error if it is another Exim or anything else that does the synchronization
5117 check. A QUIT command is no longer sent after a timeout.
5119 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5120 during host lookups.
5122 74. The defer_ok option of callout verification was not working if it was used
5123 when verifying addresses in header lines, that is, for this case:
5125 verify = header_sender/callout=defer_ok
5127 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5128 those file descriptors could be used for SMTP connections. If anything
5129 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5130 could be sent to the SMTP client, causing chaos. The daemon now opens
5131 stdin, stdout, and stderr to /dev/null when it puts itself into the
5134 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5135 log by default. The user can override this with suitable Perl magic.
5137 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5138 the log message when discard triggers, was not working for the DATA ACL or
5139 for the non-SMTP ACL.
5141 78. Error message wording change in sieve.c.
5143 79. If smtp_accept_max_per_host was set, the number of connections could be
5144 restricted to fewer than expected, because the daemon was trying to set up
5145 a new connection before checking whether the processes handling previous
5146 connections had finished. The check for completed processes is now done
5147 earlier. On busy systems, this bug wouldn't be noticed because something
5148 else would have woken the daemon, and it would have reaped the completed
5151 80. If a message was submitted locally by a user whose login name contained one
5152 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5153 It caused a spool format error. I have fixed the spool reading code. A
5154 related problem was that the "from" clause in the Received: line became
5155 illegal because of the space(s). It is now covered by ${quote_local_part.
5157 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5160 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5162 83. Patch from Sieve maintainer to fix unterminated string problem in
5163 "vacation" handling.
5165 84. Some minor changes to the Linux configuration files to help with other
5166 OS variants using glibc.
5168 85. One more patch for Sieve to update vacation handling to latest spec.
5171 ----------------------------------------------------
5172 See the note above about the 4.44 and 4.50 releases.
5173 ----------------------------------------------------
5179 1. Change 4.43/35 introduced a bug that caused file counts to be
5180 incorrectly computed when quota_filecount was set in an appendfile
5183 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5184 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5187 3. Give more explanation in the error message when the command for a transport
5188 filter fails to execute.
5190 4. There are several places where Exim runs a non-Exim command in a
5191 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5192 was being done only for the command run by the queryprogram router. It is
5193 now done for all such subprocesses. The other cases are: ${run, transport
5194 filters, and the commands run by the lmtp and pipe transports.
5196 5. Some older OS have a limit of 256 on the maximum number of file
5197 descriptors. Exim was using setrlimit() to set 1000 as a large value
5198 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5199 systems. I've change it so that if it can't get 1000, it tries for 256.
5201 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5202 was an oversight, and furthermore, ever since the addition of extra
5203 controls (e.g. 4.43/32), the checks on when to allow different forms of
5204 "control" were broken. There should now be diagnostics for all cases when a
5205 control that does not make sense is encountered.
5207 7. $recipients is now available in the predata ACL (oversight).
5209 8. Tidy the search cache before the fork to do a delivery from a message
5210 received from the command line. Otherwise the child will trigger a lookup
5211 failure and thereby defer the delivery if it tries to use (for example) a
5212 cached ldap connection that the parent has called unbind on.
5214 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5215 of $address_data from the recipient verification was clobbered by the
5216 sender verification.
5218 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5219 was its contents. (It was OK if the option was not defined at all.)
5221 11. A "Completed" log line is now written for messages that are removed from
5222 the spool by the -Mrm option.
5224 12. $host_address is now set to the target address during the checking of
5225 ignore_target_hosts.
5227 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5228 being passed; this would have caused $sender_host_name to have been used if
5229 matching the list had actually called for a host name (not very likely,
5230 since this list is usually IP addresses). A host name is now passed as
5233 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5234 code that specifies a non-restarting handler (typically sigaction() in
5235 modern systems) in an attempt to fix a rare and obscure crash bug.
5237 15. Narrowed the window for a race in the daemon that could cause it to ignore
5238 SIGCHLD signals. This is not a major problem, because they are used only to
5239 wake it up if nothing else does.
5241 16. A malformed maildirsize file could cause Exim to calculate negative values
5242 for the mailbox size or file count. Odd effects could occur as a result.
5243 The maildirsize information is now recalculated if the size or filecount
5246 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5247 support for a long time. Removed HAVE_SYS_VFS_H.
5249 18. Updated exipick to current release from John Jetmore.
5251 19. Allow an empty sender to be matched against a lookup in an address list.
5252 Previously the only cases considered were a regular expression, or an
5255 20. Exim went into a mad DNS lookup loop when doing a callout where the
5256 host was specified on the transport, if the DNS lookup yielded more than
5259 21. The RFC2047 encoding function was originally intended for short strings
5260 such as real names; it was not keeping to the 75-character limit for
5261 encoded words that the RFC imposes. It now respects the limit, and
5262 generates multiple encoded words if necessary. To be on the safe side, I
5263 have increased the buffer size for the ${rfc2047: expansion operator from
5266 22. Failure to deliver a bounce message always caused it to be frozen, even if
5267 there was an errors_to setting on the router. The errors_to setting is now
5270 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5271 canonical form (fully expanded) before being placed in
5272 $sender_host_address.
5274 24. Updated eximstats to version 1.33
5276 25. Include certificate and key file names in error message when GnuTLS fails
5277 to set them up, because the GnuTLS error message doesn't include the name
5278 of the failing file when there is a problem reading it.
5280 26. Expand error message when OpenSSL has problems setting up cert/key files.
5283 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5284 (this can affect the format of dates).
5286 28. exim_tidydb, when checking for the continued existence of a message for
5287 which it has found a message-specific retry record, was not finding
5288 messages that were in split spool directories. Consequently, it was
5289 deleting retry records that should have stayed in existence.
5291 29. eximstats updated to version 1.35
5292 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5293 1.35 - bugfix such that pie charts by volume are generated correctly
5295 30. The SPA authentication driver was not abandoning authentication and moving
5296 on to the next authenticator when an expansion was forced to fail,
5297 contradicting the general specification for all authenticators. Instead it
5298 was generating a temporary error. It now behaves as specified.
5300 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5301 (the order specifies the preference for clients). The order is now AES256,
5302 AES128, 3DES, ARCFOUR128.
5304 31. Small patch to Sieve code - explicitly set From: when generating an
5307 32. Exim crashed if a remote delivery caused a very long error message to be
5308 recorded - for instance if somebody sent an entire SpamAssassin report back
5309 as a large number of 550 error lines. This bug was coincidentally fixed by
5310 increasing the size of one of Exim's internal buffers (big_buffer) that
5311 happened as part of the Exiscan merge. However, to be on the safe side, I
5312 have made the code more robust (and fixed the comments that describe what
5315 33. Some experimental protocols are using DNS PTR records for new purposes. The
5316 keys for these records are domain names, not reversed IP addresses. The
5317 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5318 leaves it alone. Component reversal etc. now happens only for IP addresses.
5321 34. The host_aton() function is supposed to be passed a string that is known
5322 to be a valid IP address. However, in the case of IPv6 addresses, it was
5323 not checking this. This is a hostage to fortune. Exim now panics and dies
5324 if the condition is not met. A case was found where this could be provoked
5325 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5326 components; fortuitously, this particular loophole had already been fixed
5327 by change 4.50/55 or 4.44/33 above.
5329 If there are any other similar loopholes, the new check in host_aton()
5330 itself should stop them being exploited. The report I received stated that
5331 data on the command line could provoke the exploit when Exim was running as
5332 exim, but did not say which command line option was involved. All I could
5333 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5334 running as the user.
5337 35. There was a buffer overflow vulnerability in the SPA authentication code
5338 (which came originally from the Samba project). I have added a test to the
5339 spa_base64_to_bits() function which I hope fixes it.
5342 36. The daemon start-up calls getloadavg() while still root for those OS that
5343 need the first call to be done as root, but it missed one case: when
5344 deliver_queue_load_max is set with deliver_drop_privilege. This is
5345 necessary for the benefit of the queue runner, because there is no re-exec
5346 when deliver_drop_privilege is set.
5348 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5349 was in use, was not putting the data itself into the right store pool;
5350 consequently, it could be overwritten for a subsequent message in the same
5351 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5354 38. Sometimes the final signoff response after QUIT could fail to get
5355 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5356 < 0 before doing a fflush(). This bug looks as though it goes back to the
5357 introduction of TLS in release 3.20, but "sometimes" must have been rare
5358 because the tests only now provoked it.
5364 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5365 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5366 Now it returns an integer. A number of calls were still expecting a T/F
5367 return. Fortuitously, in all cases, the tests worked in OK situations,
5368 which is the norm. However, things would have gone wrong on any write
5369 failures on the smtp file descriptor. This function is used when sending
5370 messages over SMTP and also when doing verify callouts.
5372 2. When Exim is called to do synchronous delivery of a locally submitted
5373 message (the -odf or -odi options), it no longer closes stderr before doing
5376 3. Implemented the mua_wrapper option.
5378 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5380 5. Implemented the functions header_remove(), header_testname(),
5381 header_add_at_position(), and receive_remove_recipient(), and exported them
5384 6. If an ACL "warn" statement specified the addition of headers, Exim already
5385 inserted X-ACL-Warn: at the start if there was no header name. However, it
5386 was not making this test for the second and subsequent header lines if
5387 there were newlines in the string. This meant that an invalid header could
5388 be inserted if Exim was badly configured.
5390 7. Allow an ACL "warn" statement to add header lines at the start or after all
5391 the Received: headers, as well as at the end.
5393 8. Added the rcpt_4xx retry error code.
5395 9. Added postmaster_mailfrom=xxx to callout verification option.
5397 10. Added mailfrom=xxxx to the callout verification option, for verify=
5400 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5401 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5403 12. Inserted some casts to stop certain compilers warning when using pointer
5404 differences as field lengths or precisions in printf-type calls (mostly
5405 affecting debugging statements).
5407 13. Added optional readline() support for -be (dynamically loaded).
5409 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5410 same clock tick as a message's arrival, so that its received time was the
5411 same as the "first fail" time on the retry record, and that message
5412 remained on the queue past the ultimate address timeout, every queue runner
5413 would try a delivery (because it was past the ultimate address timeout) but
5414 after another failure, the ultimate address timeout, which should have then
5415 bounced the address, did not kick in. This was a "< instead of <=" error;
5416 in most cases the first failure would have been in the next clock tick
5417 after the received time, and all would be well.
5419 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5420 being recognized when the domain list was tested by the match_domain
5421 condition in an expansion string.
5423 16. Added the ${str2b64: operator.
5425 17. Exim was always calling setrlimit() to set a large limit for the number of
5426 processes, without checking whether the existing limit was already
5427 adequate. (It did check for the limit on file descriptors.) Furthermore,
5428 errors from getrlimit() and setrlimit() were being ignored. Now they are
5429 logged to the main and panic logs, but Exim does carry on, to try to do its
5430 job under whatever limits there are.
5432 18. Imported PCRE 5.0.
5434 19. Trivial typo in log message " temporarily refused connection" (the leading
5437 20. If the log selector return_path_on_delivery was set and an address was
5438 redirected to /dev/null, the delivery process crashed because it assumed
5439 that a return path would always be set for a "successful" delivery. In this
5440 case, the whole delivery is bypassed as an optimization, and therefore no
5443 21. Internal re-arrangement: the function for sending a challenge and reading
5444 a response while authentication was assuming a zero-terminated challenge
5445 string. It's now changed to take a pointer and a length, to allow for
5446 binary data in such strings.
5448 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5450 23. Exim was not respecting finduser_retries when seeking the login of the
5451 uid under which it was called; it was always trying 10 times. (The default
5452 setting of finduser_retries is zero.) Also, it was sleeping after the final
5453 failure, which is pointless.
5455 24. Implemented tls_on_connect_ports.
5457 25. Implemented acl_smtp_predata.
5459 26. If the domain in control=submission is set empty, Exim assumes that the
5460 authenticated id is a complete email address when it generates From: or
5461 Sender: header lines.
5463 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5464 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5465 chown and chgrp in /bin and hostname in /usr/bin.
5467 28. Exim was keeping the "process log" file open after each use, just as it
5468 does for the main log. This opens the possibility of it remaining open for
5469 long periods when the USR1 signal hits a daemon. Occasional processlog
5470 errors were reported, that could have been caused by this. Anyway, it seems
5471 much more sensible not to leave this file open at all, so that is what now
5474 29. The long-running daemon process does not normally write to the log once it
5475 has entered its main loop, and it closes the log before doing so. This is
5476 so that log files can straightforwardly be renamed and moved. However,
5477 there are a couple of unusual error situations where the daemon does write
5478 log entries, and I had neglected to close the log afterwards.
5480 30. The text of an SMTP error response that was received during a remote
5481 delivery was being truncated at 512 bytes. This is too short for some of
5482 the long messages that one sometimes sees. I've increased the limit to
5485 31. It is now possible to make retry rules that apply only when a message has a
5486 specific sender, in particular, an empty sender.
5488 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5489 it possible to be selective about when SMTP synchronization is enforced.
5491 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5493 32. Implemented hosts_connection_nolog.
5495 33. Added an ACL for QUIT.
5497 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5500 35. Added mailbox_size and mailbox_filecount to appendfile.
5502 36. Added control = no_multiline_responses to ACLs.
5504 37. There was a bug in the logic of the code that waits for the clock to tick
5505 in the case where the clock went backwards by a substantial amount such
5506 that the microsecond fraction of "now" was more than the microsecond
5507 fraction of "then" (but the whole seconds number was less).
5509 38. Added support for the libradius Radius client library this is found on
5510 FreeBSD (previously only the radiusclient library was supported).
5516 1. When certain lookups returned multiple values in the form name=value, the
5517 quoting of the values was not always being done properly. Specifically:
5518 (a) If the value started with a double quote, but contained no whitespace,
5520 (b) If the value contained whitespace other than a space character (i.e.
5521 tabs or newlines or carriage returns) it was not quoted.
5522 This fix has been applied to the mysql and pgsql lookups by writing a
5523 separate quoting function and calling it from the lookup code. The fix
5524 should probably also be applied to nisplus, ibase and oracle lookups, but
5525 since I cannot test any of those, I have not disturbed their existing code.
5527 2. A hit in the callout cache for a specific address caused a log line with no
5528 reason for rejecting RCPT. Now it says "Previous (cached) callout
5529 verification failure".
5531 3. There was an off-by-one bug in the queryprogram router. An over-long
5532 return line was truncated at 256 instead of 255 characters, thereby
5533 overflowing its buffer with the terminating zero. As well as fixing this, I
5534 have increased the buffer size to 1024 (and made a note to document this).
5536 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5537 when Exim is waiting for an SMTP response from a remote server, Exim
5538 restarts its select() call on the socket, thereby resetting its timeout.
5539 This is not a problem when such interrupts are rare. Somebody set up a cron
5540 job to run exiwhat every 2 minutes, which is less than the normal select()
5541 timeout (5 or 10 minutes). This meant that the select() timeout never
5542 kicked in because it was always reset. I have fixed this by comparing the
5543 time when an interrupt arrives with the time at the start of the first call
5544 to select(). If more time than the timeout has elapsed, the interrupt is
5545 treated as a timeout.
5547 5. Some internal re-factoring in preparation for the addition of Sieve
5548 extensions (by MH). In particular, the "personal" test is moved to a
5549 separate function, and given an option for scanning Cc: and Bcc: (which is
5550 not set for Exim filters).
5552 6. When Exim created an email address using the login of the caller as the
5553 local part (e.g. when creating a From: or Sender: header line), it was not
5554 quoting the local part when it contained special characters such as @.
5556 7. Installed new OpenBSD configuration files.
5558 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5559 try to make them clearer.
5561 9. Callout options, other than the timeout value, were being ignored when
5562 verifying sender addresses in header lines. For example, when using
5564 verify = header_sender/callout=no_cache
5566 the cache was (incorrectly) being used.
5568 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5569 only the Cygwin environment.
5571 11. When return_path_on_delivery was set as a log selector, if different remote
5572 addresses in the same message used different return paths and parallel
5573 remote delivery occurred, the wrong values would sometimes be logged.
5574 (Whenever a remote delivery process finished, the return path value from
5575 the most recently started remote delivery process was logged.)
5577 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5578 lines when AUTH and/or TLS are in use. This is the "received protocol"
5579 field. Exim used to use "asmtp" for authenticated SMTP, without any
5580 indication (in the protocol name) for TLS use. Now it follows the RFC and
5581 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5582 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5583 names appear in log lines as well as in Received: header lines.
5585 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5586 extensions, and comparison tests, and to fix some bugs.
5588 14. Changes to the "personal" filter test:
5590 (1) The test was buggy in that it was just doing the equivalent of
5591 "contains" tests on header lines. For example, if a user's address was
5592 anne@some.where, the "personal" test would incorrectly be true for
5594 To: susanne@some.where
5596 This test is now done by extracting each address from the header in turn,
5597 and checking the entire address. Other tests that are part of "personal"
5598 are now done using regular expressions (for example, to check local parts
5599 of addresses in From: header lines).
5601 (2) The list of non-personal local parts in From: addresses has been
5602 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5603 taken from the Sieve specification recommendations.
5605 (3) If the message contains any header line starting with "List-" it is
5606 treated as non-personal.
5608 (4) The test for "circular" in the Subject: header line has been removed
5609 because it now seems ill-conceived.
5611 15. Minor typos in src/EDITME comments corrected.
5613 16. Installed latest exipick from John Jetmore.
5615 17. If headers_add on a router specified a text string that was too long for
5616 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5617 of string_sprintf() is now avoided.
5619 18. $message_body_size was not set (it was always zero) when running the DATA
5620 ACL and the local_scan() function.
5622 19. For the "mail" command in an Exim filter, no default was being set for
5623 the once_repeat time, causing a random time value to be used if "once" was
5624 specified. (If the value happened to be <= 0, no repeat happened.) The
5625 default is now 0s, meaning "never repeat". The "vacation" command was OK
5626 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5627 (I found it when inspecting the code).
5629 20. There is now an overall timeout for performing a callout verification. It
5630 defaults to 4 times the callout timeout, which applies to individual SMTP
5631 commands during the callout. The overall timeout applies when there is more
5632 than one host that can be tried. The timeout is checked before trying the
5633 next host. This prevents very long delays if there are a large number of
5634 hosts and all are timing out (e.g. when the network connections are timing
5635 out). The value of the overall timeout can be changed by specifying an
5636 additional sub-option for "callout", called "maxwait". For example:
5638 verify = sender/callout=5s,maxwait=20s
5640 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5641 to the end before writing, but this should make it even safer).
5643 22. Exim was forgetting that it had advertised PIPELINING for the second and
5644 subsequent messages on an SMTP connection. It was also not resetting its
5645 memory on STARTTLS and an internal HELO.
5647 23. When Exim logs an SMTP synchronization error within a session, it now
5648 records whether PIPELINING has been advertised or not.
5650 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5651 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5652 rather than long int.
5654 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5656 26. Added the never_mail option to autoreply.
5662 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5663 crash if the getsockname() call failed; this can happen if a connection is
5664 closed very soon after it is established. The problem was simply in the
5665 order in which certain operations were done, causing Exim to try to write
5666 to the SMTP stream before it had set up the file descriptor. The bug has
5667 been fixed by making things happen in the correct order.
5673 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5674 before the connection was closed, thus losing the rejection response.
5676 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5677 some early Solaris releases, but causes trouble in current releases where
5678 socklen_t is defined.
5680 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5683 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5686 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5687 that releasing the top of it at the end releases what was used for sub-
5688 expansions (unless the block got too big). However, discard this block if
5689 the first thing is a variable or header, so that we can use its block when
5690 it is dynamic (useful for very large $message_headers, for example).
5692 6. Lookups now cache *every* query, not just the most recent. A new, separate
5693 store pool is used for this. It can be recovered when all lookup caches are
5694 flushed. Lookups now release memory at the end of their result strings.
5695 This has involved some general refactoring of the lookup sources.
5697 7. Some code has been added to the store_xxx() functions to reduce the amount
5698 of flapping under certain conditions.
5700 8. log_incoming_interface used to affect only the <= reception log lines. Now
5701 it causes the local interface and port to be added to several more SMTP log
5702 lines, for example "SMTP connection from", and rejection lines.
5704 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5706 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5708 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5709 could be overwritten at the end of the current message (or the start of a
5710 new message if it was set in a HELO ACL). The value is now preserved for
5711 the duration of the SMTP connection.
5713 12. If a transport had a headers_rewrite setting, and a matching header line
5714 contained an unqualified address, that address was qualified, even if it
5715 did not match any rewriting rules. The underlying bug was that the values
5716 of the flags that permit the existence of unqualified sender and recipient
5717 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5718 non-local messages, and by -bnq for local messages) were not being
5719 preserved with the message after it was received.
5721 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5722 "next input=" as part of the text comprising the host identity instead of
5723 the correct text. The code was using the same buffer for two different
5724 strings. However, depending on which order the printing function evaluated
5725 its arguments, the bug did not always show up. Under Linux, for example, my
5726 test suite worked just fine.
5728 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5729 doesn't work with some older versions of Perl. It has been changed to "my",
5730 which in any case is probably the better facility to use.
5732 15. A really picky compiler found some instances of statements for creating
5733 error messages that either had too many or two few arguments for the format
5736 16. The size of the buffer for calls to the DNS resolver has been increased
5737 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5738 for addresses that have a lot of PTR records. This alleviates a problem; it
5739 does not fully solve it.
5741 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5742 buffer now truncates the list and logs the incident, which is the same
5743 action as happens when Exim is looking up a host name and its aliases.
5744 Previously in this situation something unpredictable would happen;
5745 sometimes it was "internal error: store_reset failed".
5747 18. If a server dropped the connection unexpectedly when an Exim client was
5748 using GnuTLS and trying to read a response, the client delivery process
5749 crashed while trying to generate an error log message.
5751 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5752 string, for example:
5754 warn message = H1: something\nH2: something
5756 the text was added as a single header line from Exim's point of view
5757 though it ended up OK in the delivered message. However, searching for the
5758 second and subsequent header lines using $h_h2: did not work. This has been
5759 fixed. Similarly, if a system filter added multiple headers in this way,
5760 the routers could not see them.
5762 20. Expanded the error message when iplsearch is called with an invalid key to
5763 suggest using net-iplsearch in a host list.
5765 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5766 ACLs are no longer actually imposed (and a message to that effect is
5769 22. If a "gecos" field in a passwd entry contained escaped characters, in
5770 particular, if it contained a \" sequence, Exim got it wrong when building
5771 a From: or a Sender: header from that name. A second bug also caused
5772 incorrect handling when an unquoted " was present following a character
5773 that needed quoting.
5775 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5776 was not being matched caselessly.
5778 24. Arranged for all hyphens in the exim.8 source to be escaped with
5781 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5782 properly. Recipient callout cache records were still being keyed to include
5783 the sender, even when use_sender was set false. This led to far more
5784 callouts that were necessary. The sender is no longer included in the key
5785 when use_sender is false.
5787 26. Added "control = submission" modifier to ACLs.
5789 27. Added the ${base62d: operator to decode base 62 numbers.
5791 28. dnsdb lookups can now access SRV records.
5793 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5794 the configuration file.
5796 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5797 (-v) mode. This makes the output for a verbose queue run more intelligible.
5799 31. Added a use_postmaster feature to recipient callouts.
5801 32. Added the $body_zerocount variable, containing the number of binary zero
5802 bytes in the message body.
5804 33. The time of last modification of the "new" subdirectory is now used as the
5805 "mailbox time last read" when there is a quota error for a maildir
5808 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5810 35. Added +ignore_unknown as a special item in host lists.
5812 36. Code for decoding IPv6 addresses in host lists is now included, even if
5813 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5814 address was recognized as an IP address, but was then not correctly decoded
5815 into binary, causing unexpected and incorrect effects when compared with
5822 1. Very minor rewording of debugging text in manualroute to say "list of
5823 hosts" instead of "hostlist".
5825 2. If verify=header_syntax was set, and a header line with an unqualified
5826 address (no domain) and a large number of spaces between the end of the
5827 name and the colon was received, the reception process suffered a buffer
5828 overflow, and (when I tested it) crashed. This was caused by some obsolete
5829 code that should have been removed. The fix is to remove it!
5831 3. When running in the test harness, delay a bit after writing a bounce
5832 message to get a bit more predictability in the log output.
5834 4. Added a call to search_tidyup() just before forking a reception process. In
5835 theory, someone could use a lookup in the expansion of smtp_accept_max_
5836 per_host which, without the tidyup, could leave open a database connection.
5838 5. Added the variables $recipient_data and $sender_data which get set from a
5839 lookup success in an ACL "recipients" or "senders" condition, or a router
5840 "senders" option, similar to $domain_data and $local_part_data.
5842 6. Moved the writing of debug_print from before to after the "senders" test
5845 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5846 problems for message scanning, either using a data ACL, or using
5847 local_scan() because the Received: header was not generated till after they
5848 were called (in order to set the time as the time of reception completion).
5849 I have revised the way this works. The header is now generated after the
5850 body is received, but before the ACL or local_scan() are called. After they
5851 are run, the timestamp in the header is updated.
5857 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5858 before starting a queue runner without re-exec. This happened only when
5859 deliver_drop_privilege was set or when the Exim user was set to root. The
5860 effect of the bug was that timeouts during subsequent deliveries caused
5861 crashes instead of being properly handled. The handler is now left at its
5862 default (and expected) setting.
5864 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5865 message, again when deliver_drop_privilege is set or Exim is run as root.
5866 The bug described in (1) was not present in this case, but the tidying up
5867 of the other signals was missing. I have made the two cases consistent.
5869 3. The ignore_target_hosts setting on a manualroute router was being ignored
5870 for hosts that were looked up using the /MX notation.
5872 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5875 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5876 operated on the sender address. After changing the $sender_address to <>
5877 for the sender address verify, Exim was re-instated it as the original
5878 (before rewriting) address, but remembering that it had rewritten it, so it
5879 wasn't rewriting it again. This bug also had the effect of breaking the
5880 sender address verification caching when the sender address was rewritten.
5882 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5883 This has been changed so that if the ip literal address matches
5884 ignore_target_hosts, the router declines.
5886 7. Added expansion conditions match_domain, match_address, and match_local_
5887 part (NOT match_host).
5889 8. The placeholder for the Received: header didn't have a length field set.
5891 9. Added code to Exim itself and to exim_lock to test for a specific race
5892 condition that could lead to file corruption when using MBX delivery. The
5893 issue is with the lockfile that is created in /tmp. If this file is removed
5894 after a process has opened it but before that process has acquired a lock,
5895 there is the potential for a second process to recreate the file and also
5896 acquire a lock. This could lead to two Exim processes writing to the file
5897 at the same time. The added code performs the same test as UW imapd; it
5898 checks after acquiring the lock that its file descriptor still refers to
5899 the same named file.
5901 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5902 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5905 11. Added the smtp_active_hostname option. If used, this will typically be made
5906 to depend on the incoming interface address. Because $interface_address is
5907 not set up until the daemon has forked a reception process, error responses
5908 that can happen earlier (such as "too many connections") no longer contain
5911 12. If an expansion in a condition on a "warn" statement fails because a lookup
5912 defers, the "warn" statement is abandoned, and the next ACL statement is
5913 processed. Previously this caused the whole ACL to be aborted.
5915 13. Added the iplsearch lookup type.
5917 14. Added ident_timeout as a log selector.
5919 15. Added tls_certificate_verified as a log selector.
5921 16. Added a global option tls_require_ciphers (compare the smtp transport
5922 option of the same name). This controls incoming TLS connections.
5924 17. I finally figured out how to make tls_require_ciphers do a similar thing
5925 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5926 before starting the TLS session.
5928 18. Tabs are now shown as \t in -bP output.
5930 19. If the log selector return_path_on_delivery was set, Exim crashed when
5931 bouncing a message because it had too many Received: header lines.
5933 20. If two routers both had headers_remove settings, and the first one included
5934 a superfluous trailing colon, the final name in the first list and the
5935 first name in the second list were incorrectly joined into one item (with a
5936 colon in the middle).
5942 1. Added -C and -D options to the exinext utility, mainly to make it easier
5943 to include in the automated testing, but these could be helpful when
5944 multiple configurations are in use.
5946 2. The exinext utility was not formatting the output nicely when there was
5947 an alternate port involved in the retry record key, nor when there was a
5948 message id as well (for retries that were specific to a specific message
5949 and a specific host). It was also confused by IPv6 addresses, because of
5950 the additional colons they contain. I have fixed the IPv4 problem, and
5951 patched it up to do a reasonable job for IPv6.
5953 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5954 delivery, the log line now contains "pipelined" if PIPELINING was used.
5956 4. An SMTP transport process used to panic and die if the bind() call to set
5957 an explicit outgoing interface failed. This has been changed; it is now
5958 treated in the same way as a connect() failure.
5960 5. A reference to $sender_host_name in the part of a conditional expansion
5961 that was being skipped was still causing a DNS lookup. This no longer
5964 6. The def: expansion condition was not recognizing references to header lines
5965 that used bh_ and bheader_.
5967 7. Added the _cache feature to named lists.
5969 8. The code for checking quota_filecount in the appendfile transport was
5970 allowing one more file than it should have been.
5972 9. For compatibility with Sendmail, the command line option
5980 and sets the incoming protocol and host name (for trusted callers). The
5981 host name and its colon can be omitted when only the protocol is to be set.
5982 Note the Exim already has two private options, -pd and -ps, that refer to
5983 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5984 "s", but I don't think that's a major issue.
5986 10. A number of refactoring changes to the code, none of which should affect
5989 (a) The number of logging options was getting close to filling up the
5990 32-bit word that was used as a bit map. I have split them into two classes:
5991 those that are passed in the argument to log_write(), and those that are
5992 only ever tested independently outside of that function. These are now in
5993 separate 32-bit words, so there is plenty of room for expansion again.
5994 There is no change in the user interface or the logging behaviour.
5996 (b) When building, for example, log lines, the code previously used a
5997 macro that called string_cat() twice, in order to add two strings. This is
5998 not really sufficiently general. Furthermore, there was one instance where
5999 it was actually wrong because one of the argument was used twice, and in
6000 one call a function was used. (As it happened, calling the function twice
6001 did not affect the overall behaviour.) The macro has been replaced by a
6002 function that can join an arbitrary number of extra strings onto a growing
6005 (c) The code for expansion conditions now uses a table and a binary chop
6006 instead of a serial search (which was left over from when there were very
6007 few conditions). Also, it now recognizes conditions like "pam" even when
6008 the relevant support is not compiled in: a suitably worded error message is
6009 given if an attempt is made to use such a condition.
6011 11. Added ${time_interval:xxxxx}.
6013 12. A bug was causing one of the ddress fields not to be passed back correctly
6014 from remote delivery subprocesses. The field in question was not being
6015 subsequently used, so this caused to problems in practice.
6017 13. Added new log selectors queue_time and deliver_time.
6019 14. Might have fixed a bug in maildirsizefile handling that threw up
6020 "unexpected character" debug warnings, and recalculated the data
6021 unnecessarily. In any case, I expanded the warning message to give more
6024 15. Added the message "Restricted characters in address" to the statements in
6025 the default ACL that block characters like @ and % in local parts.
6027 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6028 Three changes have been made:
6030 (a) There was a serious bug; a negative response to MAIL caused the whole
6031 recipient domain to be cached as invalid, thereby blocking all messages
6032 to all local parts at the same domain, from all senders. This bug has
6033 been fixed. The domain is no longer cached after a negative response to
6034 MAIL if the sender used is not empty.
6036 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6039 (c) A new callout option, "use_sender" has been added for people who want
6040 the modified behaviour.
6046 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6049 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6050 indeed breaks things for older releases.
6052 3. Added additional logging to the case where there is a problem reading data
6053 from a filter that is running in a subprocess using a pipe, in order to
6054 try to track down a specific problem.
6056 4. Testing facility fudge: when running in the test harness and attempting
6057 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6058 getting "No route to host". Convert this to a timeout.
6060 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6063 6. Some OS don't have socklen_t but use size_t instead. This affects the
6064 fifth argument of getsockopt() amongst other things. This is now
6065 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6066 can be set for individual OS. I have set it for SunOS5, OSF1, and
6067 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6068 some earlier ones do not.
6070 7. Change 4.30/15 was not doing the test caselessly.
6072 8. The standard form for an IPv6 address literal was being rejected by address
6073 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6074 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6075 this, as well as the form without the "IPv6" on the front (but only when
6076 address literals are enabled, of course).
6078 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6080 10. Exim crashed if a message with an empty sender address specified by -f
6081 encountered a router with an errors_to setting. This could be provoked only
6082 by a command such as
6086 where an empty string was supplied; "<>" did not hit this bug.
6088 11. Installed PCRE release 4.5.
6090 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6091 remained set. It is now erased.
6093 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6094 times from message ids (which are base 36 rather than the normal 62).
6096 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6097 were being counted as actual protocol errors, and logged if the log
6098 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6099 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6100 and DATA following a set of rejected RCPTs do not count as protocol errors.
6101 In other words, Exim assumes they were pipelined, though this may not
6102 actually be the case. Of course, in all cases the client gets an
6103 appropriate error code.
6105 15. If a lookup fails in an ACL condition, a message about the failure may
6106 be available; it is used if testing the ACL cannot continue, because most
6107 such messages specify what the cause of the deferral is. However, some
6108 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6109 that caused an old message to be retained and used if a later statement
6110 caused a defer, replacing the real cause of the deferral.
6112 16. If an IP address had so many PTR records that the DNS lookup buffer
6113 was not large enough to hold them, Exim could crash while trying to process
6114 the truncated data. It now detects and logs this case.
6116 17. Further to 4.21/58, another change has been made: if (and only if) the
6117 first line of a message (the first header line) ends with CRLF, a bare LF
6118 in a subsequent header line has a space inserted after it, so as not to
6119 terminate the header.
6121 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6122 unnecessarily, used atoi() instead of strtol(), and didn't check the
6123 termination when getting file sizes from file names by regex.
6125 19. Completely re-implemented the support for maildirsize files, in the light
6126 of a number of problems with the previous contributed implementation
6127 (4.30/29). In particular:
6129 . If the quota is zero, the maildirsize file is maintained, but no quota is
6132 . If the maildir directory does not exist, it is created before any attempt
6133 to write a maildirsize file.
6135 . The quota value in the file is just a cache; if the quota is changed in
6136 the transport, the new value overrides.
6138 . A regular expression is available for excluding directories from the
6141 20. The autoreply transport checks the characters in options that define the
6142 message's headers; it allows continued headers, but it was checking with
6143 isspace() after an embedded newline instead of explicitly looking for a
6146 21. If all the "regular" hosts to which an address was routed had passed their
6147 expiry times, and had not reached their retry times, the address was
6148 bounced, even if fallback hosts were defined. Now Exim should go on to try
6151 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6152 equivalent code in the SMTP transport. Some hosts send humungous responses
6153 to HELO/EHLO, more than 1024 it seems.
6155 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6156 gives compiler warnings in some environments. I've now done it "properly",
6159 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6160 (because of problems with the built-in one) was declared to return uschar *
6161 instead of char *, causing compiler failure.
6163 25. Fixed a file descriptor leak when processing alias/forward files.
6165 26. Fixed a minor format string issue in dbfn.c.
6167 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6169 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6170 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6173 29. When a sender address is verified, it is cached, to save repeating the test
6174 when there is more than one recipient in a message. However, when the
6175 verification involves a callout, it is possible for different callout
6176 options to be set for different recipients. It is too complicated to keep
6177 track of this in the cache, so now Exim always runs a verification when a
6178 callout is required, relying on the callout cache for the optimization.
6179 The overhead is duplication of the address routing, but this should not be
6182 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6183 to be verified with callout=postmaster, and the main callout worked but the
6184 postmaster check failed, the verification correctly failed. However, if a
6185 subsequent RCPT command asked for sender verification *without* the
6186 postmaster check, incorrect caching caused this verification also to fail,
6189 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6190 it was not caching the DNS options (qualify_single, search_parents) that
6191 were used when the lookup failed. A subsequent lookup with different
6192 options therefore always gave the same answer, though there were cases
6193 where it should not have. (Example: a "domains = !$mx_any" option on a
6194 dnslookup router: the "domains" option is always processed without any
6195 widening, but the router might have qualify_single set.) Now Exim uses the
6196 cached value only when the same options are set.
6198 32. Added John Jetmore's "exipick" utility to the distribution.
6200 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6201 than a timeout (e.g. a certificate is required, and is not provided), an
6202 Exim server now closes the connection immediately. Previously it waited for
6203 the client to close - but if the client is SSL, it seems that they each
6204 wait for each other, leading to a delay before one of them times out.
6206 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6207 maintained 0.8.x compatibility because I don't think many are using it, and
6208 it is clearly obsolete.
6210 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6213 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6214 client certificate was expired. A simple patch fixes this, though I don't
6215 understand the full logic of why the verify callback is called multiple
6218 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6219 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6220 which causes problems with some clients (such as the Certicom SSL Plus
6221 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6222 disables the coutermeasure allowing Eudora to connect."
6224 38. Exim was not checking that a write() to a log file succeeded. This could
6225 lead to Bad Things if a log got too big, in particular if it hit a file
6226 size limit. Exim now panics and dies if it cannot write to a log file, just
6227 as it does if it cannot open a log file.
6229 39. Modified OS/Makefile-Linux so that it now contains
6231 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6233 The two -D definitions ensure that Exim is compiled with large file
6234 support, which makes it possible to handle log files that are bigger than
6237 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6238 instance) a domain was checked against a named list that involved a lookup,
6239 causing $domain_data to be set, then another domain was checked against the
6240 same list, then the first domain was re-checked, the value of $domain_data
6241 after the final check could be wrong. In particular, if the second check
6242 failed, it could be set empty. This bug probably also applied to
6245 41. The strip_trailing_dot option was not being applied to the address given
6246 with the -f command-line option.
6248 42. The code for reading a message's header from the spool was incrementing
6249 $received_count, but never initializing it. This meant that the value was
6250 incorrect (doubled) while delivering a message in the same process in which
6251 it was received. In the most common configuration of Exim, this never
6252 happens - a fresh exec is done - but it can happen when
6253 deliver_drop_privilege is set.
6255 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6256 it now includes up to 150 characters of the unexpected data in the log
6259 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6260 and building data strings. The size of both of these buffers was 10 000
6261 bytes - far larger than anybody would *ever* want, thought I. Needless to
6262 say, somebody hit the limit. I have increased the maximum line length to
6263 20 000 and the maximum data length of concatenated lines to 100 000. I have
6264 also fixed two bugs, because there was no checking on these buffers. Tsk,
6265 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6266 buffer is too small.
6268 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6269 lsearch lookups. Now it does.
6271 46. When parsing a route_list item in a manualroute router, a fixed-length
6272 buffer was used for the list of hosts. I made this 1024 bytes long,
6273 thinking that nobody would ever have a list of hosts that long. Wrong.
6274 Somebody had a whole pile of complicated expansion conditions, and the
6275 string was silently truncated, leading to an expansion error. It turns out
6276 that it is easier to change to an unlimited length (owing to other changes
6277 that have happened since this code was originally written) than to build
6278 structure for giving a limitation error. The length of the item that
6279 expands into the list of hosts is now unlimited.
6281 47. The lsearch lookup could not handle data where the length of text line was
6282 more than 4095 characters. Such lines were truncated, leading to shortened
6283 data being returned. It should now handle lines of any length.
6285 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6286 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6289 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6290 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6291 no longer happens in comment lines. A list of replacements is now placed
6292 at the head of all of the source files, except those whose only change is
6293 to replace PERL_COMMAND in the very first #! line.
6295 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6296 messages on the queue, with a bottom-up merge sort, using code contributed
6297 by Michael Haardt. This should make operations like -bp somewhat faster on
6298 large queues. It won't affect queue runners, except when queue_run_in_order
6301 51. Installed eximstats 1.31 in the distribution.
6303 52. Added support for SRV lookups to the dnslookup router.
6305 53. If an ACL referred to $message_body or $message_body_end, the value was not
6306 reset for any messages that followed in the same SMTP session.
6308 54. The store-handling optimization for building very long strings was not
6309 differentiating between the different store pools. I don't think this
6310 actually made any difference in practice, but I've tidied it.
6312 55. While running the routers to verify a sender address, $sender_address
6313 was still set to the sender address. This is wrong, because when routing to
6314 send a bounce to the sender, it would be empty. Therefore, I have changed
6315 it so that, while verifying a sender address, $sender_address is set to <>.
6316 (There is no change to what happens when verifying a recipient address.)
6318 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6319 target A or AAAA records (if not already returned) without resetting the
6320 qualify_single or search_parents options of the DNS resolver. These are
6321 inappropriate in this case because the targets of MX and SRV records must
6322 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6323 target that, when qualified, matched something in the local domain. These
6324 two options are now turned off when doing these lookups.
6326 57. It seems that at least some releases of Reiserfs (which does not have the
6327 concept of a fixed number of inodes) returns zero and not -1 for the
6328 number of available inodes. This interacted badly with check_spool_inodes,
6329 which assumed that -1 was the "no such thing" setting. What I have done is
6330 to check that the total number of inodes is greater than zero before doing
6331 the test of how many are available.
6333 58. When a "warn" ACL statement has a log_message modifier, the message is
6334 remembered, and not repeated. This is to avoid a lot of repetition when a
6335 message has many recipients that cause the same warning to be written.
6336 However, Exim was preserving the list of already written lines for an
6337 entire SMTP session, which doesn't seem right. The memory is now reset if a
6338 new message is started.
6340 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6341 debugging output unless log_rewrite was also set.
6343 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6344 of (int)(handle) when we know that handle contains (void *)(-1).
6346 61. The Exim daemon panic-logs an error return when it closes the incoming
6347 connection. However "connection reset by peer" seems to be common, and
6348 isn't really an error worthy of noting specially, so that particular error
6351 62. When Exim is trying to find all the local interfaces, it used to panic and
6352 die if the ioctl to get the interface flags failed. However, it seems that
6353 on at least one OS (Solaris 9) it is possible to have an interface that is
6354 included in the list of interfaces, but for which you get a failure error
6355 for this call. This happens when the interface is not "plumbed" into a
6356 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6357 failure of the "get flags" call assumes that the interface is down.
6359 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6360 makes life easier for people who are doing arithmetic on fields extracted
6361 from dates, where you often get leading zeros that should not be
6362 interpreted as octal.
6364 64. Added qualify_domain to the redirect router, to override the global
6367 65. If a pathologically long header line contained very many addresses (the
6368 report of this problem mentioned 10 000) and each of them was rewritten,
6369 Exim could use up a very large amount of memory. (It kept on making new
6370 copies of the header line as it rewrote, and never released the old ones.)
6371 At the expense of a bit more processing, the header rewriting function has
6372 been changed so that it no longer eats memory in this way.
6374 66. The generation of the Received: header has been moved from the time that a
6375 message starts to be received, to the time that it finishes. The timestamp
6376 in the Received: header should now be very close to that of the <= log
6377 line. There are two side-effects of this change:
6379 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6380 logged header lines no longer include the local Received: line, because
6381 it has not yet been created. The same applies to a copy of the message
6382 that is returned to a non-SMTP sender when a message is rejected.
6384 (b) When a filter file is tested using -bf, no additional Received: header
6385 is added to the test message. After some thought, I decided that this
6388 This change does not affect the value of $received_for. It is still set
6389 after address rewriting, but before local_scan() is called.
6391 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6393 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6394 gave an unhelpful panic error message, and a defer error. I have managed to
6395 change this behaviour so that it now rejects any supplied certificate,
6396 which seems right, as the list of acceptable certificates is empty.
6398 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6399 gave an unhelpful defer error. I have not managed to make this reject any
6400 supplied certificates, but the error message it gives is "no certificate
6401 supplied", which is not helpful.
6403 70. exigrep's output now also includes lines that are not associated with any
6404 message, but which match the given pattern. Implemented by a patch from
6405 Martin Sluka, which also tidied up the Perl a bit.
6407 71. Recipient callout verification, like sender verification, was using <> in
6408 the MAIL FROM command. This isn't really the right thing, since the actual
6409 sender may affect whether the remote host accepts the recipient or not. I
6410 have changed it to use the actual sender in the callout; this means that
6411 the cache record is now keyed on a recipient/sender pair, not just the
6412 recipient address. There doesn't seem to be a real danger of callout loops,
6413 since a callout by the remote host to check the sender would use <>.
6414 [SEE ABOVE: changed after hitting problems.]
6416 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6417 temporary errors. However, in the case of such a code being given after
6418 the end of a data transmission (i.e. after ".") Exim was failing to write
6419 a retry record for the message. (Yes, there was some broken host that was
6420 actually sending 8xx at this point.)
6422 73. An unknown lookup type in a host list could cause Exim to panic-die when
6423 the list was checked. (An example that provoked this was putting <; in the
6424 middle of a list instead of at the start.) If this happened during a DATA
6425 ACL check, a -D file could be left lying around. This kind of configuration
6426 error no longer causes Exim to die; instead it causes a defer error. The
6427 incident is still logged to the main and panic logs.
6429 74. Buglet left over from Exim 3 conversion. The message "too many messages
6430 in one connection" was written to the rejectlog but not the mainlog, except
6431 when address rewriting (yes!) was being logged.
6433 75. Added write_rejectlog option.
6435 76. When a system filter was run not as root (that is, when system_filter_user
6436 was set), the values of the $n variables were not being returned to the
6437 main process; thus, they were not subsequently available in the $sn
6440 77. Added +return_path_on_delivery log selector.
6442 78. A connection timeout was being treated differently from recipients deferred
6443 when testing hosts_max_try with a message that was older than the host's
6444 retry timeout. (The host should not be counted, thus allowing all hosts to
6445 be tried at least once before bouncing.) This may have been the cause of an
6446 occasionally reported bug whereby a message would remain on the queue
6447 longer than the retry timeout, but would be bounced if a delivery was
6448 forced. I say "may" because I never totally pinned down the problem;
6449 setting up timeout/retry tests is difficult. See also the next item.
6451 79. The ultimate address timeout was not being applied to errors that involved
6452 a combination of host plus message (for example, a timeout on a MAIL
6453 command). When an address resolved to a number of possible hosts, and they
6454 were not all tried for each delivery (e.g. because of hosts_max_try), a
6455 message could remain on the queue longer than the retry timeout.
6457 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6460 81. Fixed an obscure SMTP outgoing bug which required at least the following
6461 conditions: (a) there was another message waiting for the same server;
6462 (b) the server returned 5xx to all RCPT commands in the first message so
6463 that the message was not completed; (c) the server dropped the connection
6464 or gave a negative response to the RSET that Exim sends to abort the
6465 transaction. The observed case was a dropped connection after DATA that had
6466 been sent in pipelining mode. That is, the server had advertised PIPELINING
6467 but was not implementing it correctly. The effect of the bug was incorrect
6468 behaviour, such as trying another host, and this could lead to a crash.
6474 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6475 and daemon.c were passed as pointers to ints; they should have been
6476 pointers to socklen_t variables (which are typically unsigned ints).
6478 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6481 3. Fixed a really odd bug that affected only the testing scheme; patching a
6482 certain fixed string in the binary changed the value of another string that
6483 happened to be identical to the end of the original first string.
6485 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6486 name", it returns that address as the IP address. On some operating
6487 systems (e.g. Solaris), it also passes back the IP address string as the
6488 "host name". However, on others (e.g. Linux), it passes back an empty
6489 string. Exim wasn't checking for this, and was changing the host name to an
6490 empty string, assuming it had been canonicalized.
6492 5. Although rare, it is permitted to have more than one PTR record for a given
6493 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6494 all the names associated with an address, because they do in Solaris.
6495 However, it seems that they do not in Linux for data that comes from the
6496 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6497 I found this out when I moved to a new Linux workstation and tried to run
6498 the Exim test suite.
6500 To get round this problem I have changed the code so that it now does its
6501 own call to the DNS to look up PTR records when searching for a host name.
6502 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6503 addresses that are only in /etc/hosts are still found.
6505 This behaviour is, however, controlled by an option called host_lookup_
6506 order, which defaults to "bydns:byaddr". If people want to use the other
6507 order, or indeed, just use one or the other means of lookup, they can
6508 specify it in this variable.
6510 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6511 some operating systems, this comes back from gethostbyaddr() as an empty
6512 string, and this is what Exim used to test for. However, it seems that in
6513 other systems, "." is yielded. Exim now tests for this case too.
6515 7. The values of check_spool_space and check_log_space are now held internally
6516 as a number of kilobytes instead of an absolute number of bytes. If a
6517 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6518 kilobyte. This means that much larger values can be stored.
6520 8. Exim monitor: an attempt to get the action menu when not actually pointing
6521 at a message produces an empty menu entitled "No message selected". This
6522 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6523 no entries in it ("Shell widget menu has zero width and/or height"). So I
6524 have added a single, blank menu entry in this case.
6526 9. Added ${quote_local_part.
6528 10. MIME decoding is now applied to the contents of Subject: header lines when
6531 11. Now that a reference to $sender_host_address automatically causes a reverse
6532 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6533 host lookup before query-style lookups in lists that might use this
6534 variable. This has therefore been abolished, and the "net-" prefix is no
6535 longer necessary for query-style lookups.
6537 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6538 have been a typo for LFLAGS, so it has been changed.
6540 13. The install script calls Exim with "-C /dev/null" in order to find the
6541 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6542 to be output. However, since Exim outputs its version number before the
6543 error, it didn't break the script. It just looked ugly. I fixed this by
6544 always allowing "-C /dev/null" if the caller is root.
6546 14. Ignore overlarge ACL variable number when reading spool file - insurance
6547 against a later release with more variables having written the file.
6549 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6550 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6551 this, as well as the form without the "IPv6" on the front.
6553 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6554 OS/Makefile-Darwin file.
6556 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6557 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6559 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6560 variable. However, some OS use ints for inodes. I've added cast to long int
6561 to get rid of the compiler warning.
6563 19. I had forgotten to lock out "/../" in configuration file names when
6564 ALT_CONFIG_PREFIX was set.
6566 20. Routers used for verification do not need to specify transports. However,
6567 if such a router generated a host list, and callout was configured, Exim
6568 crashed, because it could not find a port number from the (non-existent)
6569 transport. It now assumes port 25 in this circumstance.
6571 21. Added the -t option to exigrep.
6573 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6574 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6575 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6576 them is set, LOOKUP_LSEARCH is forced.
6578 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6580 24. Added sender and host information to the "rejected by local_scan()" log
6581 line; previously there was no indication of these.
6583 25. Added .include_if_exists.
6585 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6586 Linux. It turns out that not all file systems support this. Apparently some
6587 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6588 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6589 not supported on the file descriptor, is now ignored when Exim is trying to
6590 sync a directory. This applies only to Linux.
6592 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6594 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6595 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6598 29. Added support for maildirsize files from supplied patch (modified a bit).
6600 30. The use of :fail: followed by an empty string could lead Exim to respond to
6601 sender verification failures with (e.g.):
6603 550 Verification failed for <xxx>
6604 550 Sender verify failed
6606 where the first response line was missing the '-' that indicates it is not
6607 the final line of the response.
6609 31. The loop for finding the name of the user that called Exim had a hardwired
6610 limit of 10; it now uses the value of finduser_retries, which is used for
6611 all other user lookups.
6613 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6616 33. Exim was neglecting to zero errno before one call of strtol() when
6617 expanding a string and expecting an integer value. On some systems this
6618 resulted in spurious "integer overflow" errors. Also, it was casting the
6619 result into an int without checking.
6621 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6622 did not work. The code looks as if it has *never* worked, though it appears
6623 to have been documented since at least release 1.62. I have made it work.
6625 35. The "timeout_DNS" error in retry rules, also documented since at least
6626 1.62, also never worked. As it isn't clear exactly what this means, and
6627 clearly it isn't a major issue, I have abolished the feature by treating it
6628 as "timeout", and writing a warning to the main and panic logs.
6630 36. The display of retry rules for -brt wasn't always showing the error code
6633 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6634 timeout_connect_A, timeout_connect_MX.
6636 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6637 to the empty sender.
6639 39. The daemon was not analysing the content of -oX till after it had closed
6640 stderr and disconnected from the controlling terminal. This meant that any
6641 syntax errors were only noted on the panic log, and the return code from
6642 the command was 0. By re-arranging the code a little, I've made the
6643 decoding happen first, so such errors now appear on stderr, and the return
6644 code is 1. However, the actual setting up of the sockets still happens in
6645 the disconnected process, so errors there are still only recorded on the
6648 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6649 connections (as happens on some IP stacks) was logged at start up time as
6650 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6651 it from "IPv6 and IPv4", which means that two separate sockets are being
6654 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6655 says whether AF_INET or AF_INET6 was passed as an argument.
6657 42. Exiwhat output was messed up when time zones were included in log
6660 43. Exiwhat now gives more information about the daemon's listening ports,
6661 and whether -tls-on-connect was used.
6663 44. The "port" option of the smtp transport is now expanded.
6665 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6666 silently ignored. Now an error message is written to the main and panic
6669 46. There's a new ACL modifier called "logwrite" which writes to a log file
6670 as soon as it is encountered.
6672 47. Added $local_user_uid and $local_user_gid at routing time.
6674 48. Exim crashed when trying to verify a sender address that was being
6677 49. Exim was recognizing only a space character after ".include". It now also
6678 recognizes a tab character.
6680 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6681 extracting the relevant information from the specification. The man page no
6682 longer contains scrambled data for the -d option, and I've added a section
6683 at the front about calling Exim under different names.
6685 51. Added "extra_headers" argument to the "mail" command in filter files.
6687 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6690 53. Installed eximstats 1.29.
6692 54. Added transport_filter_timeout as a generic transport option.
6694 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6695 Cc: header lines. This was required by RFC 822, but it not required by RFC
6698 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6699 incoming messages that did not have them. Now it does so only if the
6700 message originates locally, that is, if there is no associated remote host
6701 address. When Resent- header lines are present, this applies to the Resent-
6702 lines rather than the non-Resent- lines.
6704 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6705 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6707 58. Messages for configuration errors now include the name of the main
6708 configuration file - useful now that there may be more than one file in a
6709 list (.included file names were always shown).
6711 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6712 for those rare installations that do not start the daemon as root or run it
6713 setuid root. I've cut out the call to initgroups() if the daemon is not
6716 60. The Exim user and group can now be bound into the binary as text strings
6717 that are looked up at the start of Exim's processing.
6719 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6721 62. Added $mailstore_basename variable.
6723 63. Installed patch to sieve.c from Michael Haardt.
6725 64. When Exim failed to open the panic log after failing to open the main log,
6726 the original message it was trying to log was written to stderr and debug
6727 output, but if they were not available (the usual case in production), it
6728 was lost. Now it is written to syslog before the two lines that record the
6729 failures to open the logs.
6731 65. Users' Exim filters run in subprocesses under the user's uid. It is
6732 possible for a "deliver" command or an alias in a "personal" command to
6733 provoke an address rewrite. If logging of address rewriting is configured,
6734 this fails because the process is not running as root or exim. There may be
6735 a better way of dealing with this, but for the moment (because 4.30 needs
6736 to be released), I have disabled address rewrite logging when running a
6737 filter in a non-root, non-exim process.
6743 1. The buildconfig auxiliary program wasn't quoting the value set for
6744 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6745 not defined. This bug was masked in 4.22 by the effect that was fixed in
6748 2. Some messages that were rejected after a message id was allocated were
6749 shown as "incomplete" by exigrep. It no longer does this for messages that
6750 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6752 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6753 have allow_domain_literals set, the ID did not get logged in the <= line.
6754 Domain literals are now always recognized in Message-ID: header lines.
6756 4. The first argument for a ${extract expansion item is the key name or field
6757 number. Leading and trailing spaces in this item were not being ignored,
6758 causing some misleading effects.
6760 5. When deliver_drop_privilege was set, single queue runner processes started
6761 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6762 same command in the process it spins off) were not dropping privilege.
6764 6. When the daemon running as "exim" started a queue runner, it always
6765 re-executed Exim in the spun-off process. This is a waste of effort when
6766 deliver_drop_privilege is set. The new process now just calls the
6767 queue-runner function directly.
6773 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6776 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6777 ignored. Though the use of -oP was forcing the writing of a pid file, it
6778 was always written to the default place.
6780 3. If the message "no IP address found for host xxxx" is generated during
6781 incoming verification, it is now followed by identification of the incoming
6782 connection (so you can more easily find what provoked it).
6784 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6786 5. Added some features to "harden" Exim a bit more against certain attacks:
6788 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6789 be put in Local/Makefile. This is like the never_users runtime option,
6790 but it cannot be overridden. The default setting is "root".
6792 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6793 prefix string with which any file named in a -C command line option
6796 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6797 is retained for -C and -D only if the caller of Exim is root. Without
6798 it, the exim user may also use -C and -D and retain privilege.
6800 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6801 command line option is disabled.
6803 6. Macro names set by the -D option must start with an upper case letter, just
6804 like macro names defined in the configuration file.
6806 7. Added "dereference=" facility to LDAP.
6808 8. Two instances of the typo "uknown" in the source files are fixed.
6810 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6811 the Configure-Makefile script screwed up while processing it.
6813 10. Incorporated PCRE 4.4.
6815 11. The SMTP synchronization check was not operating right at the start of an
6816 SMTP session. For example, it could not catch a HELO sent before the client
6817 waited for the greeting. There is now a check for outstanding input at the
6818 point when the greeting is written. Because of the duplex, asynchronous
6819 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6820 way, but not yet received, when the check is performed.
6822 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6823 on TCP/IP sockets, because this apparently causes some broken clients to
6826 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6827 unchanged) from the Cygwin maintainer.
6829 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6830 was supported instead of testing for mbx. Effectively a typo.
6832 15. The spa authenticator server code was not checking that the input it
6833 received was valid base64.
6835 16. The debug output line for the "set" modifier in ACLs was not showing the
6836 name of the variable that was being set.
6838 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6840 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6841 lookup on its own. Something else was needed to trigger the lookup. For
6842 example, a match in host_lookup or the need for a host name in a host list.
6843 Now, if $sender_host_name is referenced and the host name has not yet been
6844 looked up, a lookup is performed. If the lookup fails, the variable remains
6845 empty, and $host_lookup_failed is set to "1".
6847 19. Added "eqi" as a case-independent comparison operator.
6849 20. The saslauthd authentication condition could segfault if neither service
6850 nor realm was specified.
6852 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6853 error message that was logged was misleading, and incoming SMTP
6854 connections were dropped. The message is now more accurate, and temporary
6855 errors are given to SMTP connections.
6857 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6858 (except RSET) with a 421 error, until QUIT is received. However, it was
6859 failing to send a response to QUIT.
6861 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6862 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6863 ACL is now run after the helo_try_verify_hosts code.
6865 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6866 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6867 case-independent, so other case variants are also recognized). Apparently
6868 some systems use these upper case variants.
6870 25. If more than two messages were waiting for the same host, and a transport
6871 filter was specified for the transport, Exim sent two messages over the
6872 same TCP/IP connection, and then failed with "socket operation on non-
6873 socket" when it tried to send the third.
6875 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6877 27. The extern definition of crypt16() in expand.c was not being excluded when
6878 the OS had its own crypt16() function.
6880 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6881 as a preferred synonym for return_size_limit, both as an option and as an
6884 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6886 30. Changed the default configuration ACL to relax the local part checking rule
6887 for addresses that are not in any local domains. For these addresses,
6888 slashes and pipe symbols are allowed within local parts, but the sequence
6889 /../ is explicitly forbidden.
6891 31. SPA server authentication was not clearing the challenge buffer before
6894 32. log_message in a "warn" ACL statement was writing to the reject log as
6895 well as to the main log, which contradicts the documentation and doesn't
6896 seem right (because no rejection is happening). So I have stopped it.
6898 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6899 However, I am unable to do any testing of this.
6901 34. Fixed an infelicity in the appendfile transport. When checking directories
6902 for a mailbox, to see if any needed to be created, it was accidentally
6903 using path names with one or more superfluous leading slashes; tracing
6904 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6906 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6907 contents are added to the log line that is written for every discarded
6908 recipient. (Previously a log_message setting was ignored.)
6910 36. The ${quote: operator now quotes the string if it is empty.
6912 37. The install script runs exim in order to find its version number. If for
6913 some reason other than non-existence or emptiness, which it checks, it
6914 could not run './exim', it was installing it with an empty version number,
6915 i.e. as "exim-". This error state is now caught, and the installation is
6918 38. An argument was missing from the function that creates an error message
6919 when Exim fails to connect to the socket for saslauthd authentication.
6920 This could cause Exim to crash, or give a corrupted message.
6922 39. Added isip, isip4, and isip6 to ${if conditions.
6924 40. The ACL variables $acl_xx are now saved with the message, and can be
6925 accessed later in routers, transports, and filters.
6927 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6928 strings in the file are not string-expanded.
6930 42. If a MAIL command specified a SIZE value that was too large to fit into an
6931 int variable, the check against message_size_limit failed. Such values are
6932 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6933 day this will have to be increased, but I don't think I want to be around
6934 when emails are that large.
6941 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6942 iconv() is not standard in FreeBSD.
6944 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6945 IPv6 enabled. The observed symptom was a segmentation fault on return from
6946 the function os_common_find_running_interfaces() in src/os.c.
6948 3. In the check_special_case() function in daemon.c I had used "errno" as an
6949 argument name, which causes warnings on some systems. This was basically a
6950 typo, since it was named "eno" in the comments!
6952 4. The code that waits for the clock to tick (at a resolution of some fraction
6953 of a second) so as to ensure message-id uniqueness was always waiting for
6954 at least one whole tick, when it could have waited for less. [This is
6955 almost certainly not relevant at current processor speeds, where it is
6956 unlikely to ever wait at all. But we try to future-proof.]
6958 5. The function that sleeps for a time interval that includes fractions of a
6959 second contained a race. It did not block SIGALRM between setting the
6960 timer, and suspending (a couple of lines later). If the interval was short
6961 and the sigsuspend() was delayed until after it had expired, the suspension
6962 never ended. On busy systems this could lead to processes getting stuck for
6965 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6966 process, before it forks any delivery processes. The open lookup caching
6967 mechanism meant that the open file or database connection was passed into
6968 the delivery process. The problem was that delivery processes always tidy
6969 up cached lookup data. This could cause a problem for the next delivery
6970 process started by the queue runner, because the external queue runner
6971 process does not know about the closure. So the next delivery process
6972 still has data in the lookup cache. In the case of a file lookup, there was
6973 no problem because closing a file descriptor in a subprocess doesn't affect
6974 the parent. However, if the lookup was caching a connection to a database,
6975 the connection was closed, and the second delivery process was likely to
6976 see errors such as "PGSQL: query failed: server closed the connection
6977 unexpectedly". The problem has been fixed by closing all cached lookups
6978 in a queue runner before running a delivery process.
6980 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6981 seem to have the "const" qualifier which it has on other OS. I've
6984 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6985 *of the same type* (client or server) with the same public name that an
6986 error should be diagnosed.
6988 9. When Exim looked up a host name for an IP address, but failed to find the
6989 original IP address when looking up the host name (a safety check), it
6990 output the message "<ip address> does not match any IP for NULL", which was
6991 confusing, to say the least. The bug was that the host name should have
6992 appeared instead of "NULL".
6994 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6995 user that is built into the binary, and the -C or -D options is used, root
6996 privilege is dropped before the configuration file is read. In addition,
6997 logging is switched to stderr instead of the normal log files. If the
6998 configuration then re-defines the Exim user, the unprivileged environment
6999 is probably not what is expected, so Exim logs a panic warning message (but
7002 However, if deliver_drop_privilege is set, the unprivileged state may well
7003 be exactly what is intended, so the warning has been cut out in that case,
7004 and Exim is allowed to try to write to its normal log files.
7010 1. smtp_return_error_details was not giving details for temporary sender
7011 or receiver verification errors.
7013 2. Diagnose a configuration error if two authenticators have the same public
7016 3. Exim used not to create the message log file for a message until the first
7017 delivery attempt. This could be confusing when incoming messages were held
7018 for policy or load reasons. The message log file is now created at the time
7019 the message is received, and an initial "Received" line is written to it.
7021 4. The automatically generated man page for command line options had a minor
7022 bug that caused no ill effects; however, a more serious problem was that
7023 the procedure for building the man page automatically didn't always
7024 operate. Consequently, release 4.20 contains an out-of-date version. This
7025 shouldn't happen again.
7027 5. When building Exim with embedded Perl support, the script that builds the
7028 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7029 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7031 6. The freeze_tell option was not being used for messages that were frozen on
7032 arrival, either by an ACL or by local_scan().
7034 7. Added the smtp_incomplete_transaction log selector.
7036 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7037 was accepting AUTH without a new EHLO.
7039 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7040 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7043 10. Logging of TCP/IP connections (when configured) now happens in the main
7044 daemon process instead of the child process, so that the TCP/IP connection
7045 count is more accurate (but it can never be perfect).
7047 11. The use of "drop" in a nested ACL was not being handled correctly in the
7048 outer ACL. Now, if condition failure induced by the nested "drop" causes
7049 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7050 or "require"), the connection is dropped.
7052 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7053 that yield "discard" can now be used with an "accept" or a "discard" verb,
7054 but an error is generated for any others (because I can't see a useful way
7055 to define what should happen).
7057 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7058 are now processed in the same way as lines in the Exim configuration file.
7059 In particular, continuation lines are supported.
7061 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7063 15. Added -ti meaning -t -i.
7065 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7066 domains, and warn by logging if others are found.
7068 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7069 returned from GIFCONF when Exim is trying to find the list of interfaces on
7070 a host. The code in os.c has been modified to copy each ifreq to an aligned
7071 structure in all cases.
7073 Also, in some cases, the returned ifreq's were being copied to a 'struct
7074 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7075 means the last couple of bytes of an IPv6 address could be chopped if the
7076 ifreq contained only a normal sockaddr (14 bytes storage).
7078 18. Named domain lists were not supported in the hosts_treat_as_local option.
7079 An entry such as +xxxx was not recognized, and was treated as a literal
7082 19. Ensure that header lines added by a DATA ACL are included in the reject log
7083 if the ACL subsequently rejects the message.
7085 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7086 MD5 (which is deprecated).
7088 21. When testing a filter file using -bf, Exim was writing a message when it
7089 took the sender from a "From " line in the message, but it was not doing so
7090 when it took $return_path from a Return-Path: header line. It now does.
7092 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7093 with a valid header line field name (a series of printing characters
7094 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7096 23. Changed "disc" in the source to "disk" to conform to the documentation and
7097 the book and for uniformity.
7099 24. Ignore Sendmail's -Ooption=value command line item.
7101 25. When execve() failed while trying to run a command in a pipe transport,
7102 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7103 could be confused with a return value of 69 from the command itself. This
7104 has been changed to 127, the value the shell returns if it is asked to run
7105 a non-existent command. The wording for the related log line suggests a
7106 non-existent command as the problem.
7108 26. If received_header_text expands to an empty string, do not add a Received:
7109 header line to the message. (Well, it adds a token one on the spool, but
7110 marks it "old" so that it doesn't get used or transmitted.)
7112 27. Installed eximstats 1.28 (addition of -nt option).
7114 28. There was no check for failure on the call to getsockname() in the daemon
7115 code. This can fail if there is a shortage of resources on the system, with
7116 ENOMEM, for example. A temporary error is now given on failure.
7118 29. Contrary to the C standard, it seems that in some environments, the
7119 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7120 program. Exim now does this explicitly; it affects the formatting of
7121 timestamps using strftime().
7123 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7124 complaints. I've now initialized all the variables, to avoid this.
7126 32. Header lines added by a system filter were not being "seen" during
7127 transport-time rewrites.
7129 33. The info_callback() function passed to OpenSSL is set up with type void
7130 (*)(SSL *, int, int), as described somewhere. However, when calling the
7131 function (actually a macro) that sets it up, the type void(*)() is
7132 expected. I've put in a cast to prevent warnings from picky compilers.
7134 34. If a DNS black list lookup found a CNAME record, but there were no A
7135 records associated with the domain it pointed at, Exim crashed.
7137 35. If a DNS black list lookup returned more than one A record, Exim ignored
7138 all but the first. It now scans all returned addresses if a particular IP
7139 value is being sought. In this situation, the contents of the
7140 $dnslist_value variable are a list of all the addresses, separated by a
7143 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7144 to accept a host name and all its aliases if the forward lookup for any of
7145 them yielded the IP address of the incoming connection. Now it accepts only
7146 those names whose forward lookup yields the correct IP address. Any other
7147 names are discarded. This closes a loophole whereby a rogue DNS
7148 administrator could create reverse DNS records to break through a
7149 wildcarded host restriction in an ACL.
7151 37. If a user filter or a system filter that ran in a subprocess used any of
7152 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7153 the wrong values were passed to the pipe command ($thisaddress had the
7154 value of $0, $0 had the value of $1, etc). This bug was introduced by
7155 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7157 38. Improved the line breaking for long SMTP error messages from ACLs.
7158 Previously, if there was no break point between 40 and 75 characters, Exim
7159 left the rest of the message alone. Two changes have been made: (a) I've
7160 reduced the minimum length to 35 characters; (b) if it can't find a break
7161 point between 35 and 75 characters, it looks ahead and uses the first one
7162 that it finds. This may give the occasional overlong line, but at least the
7163 remaining text gets split now.
7165 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7166 file descriptors might be low, and that setting 1000 would always raise it.
7167 It turns out that in some environments, the limit is already over 1000 and
7168 that lowering it causes trouble. So now Exim takes care not to decrease it.
7170 40. When delivering a message, the value of $return_path is set to $sender_
7171 address at the start of routing (routers may change the value). By an
7172 oversight, this default was not being set up when an address was tested by
7173 -bt or -bv, which affected the outcome if any router or filter referred to
7176 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7177 log only when "message" or "log_message" are set. However, if one of the
7178 conditions was an address verification, or a call to a nested ACL, the
7179 messages generated by the underlying test were being passed through. This
7180 no longer happens. The underlying message is available in $acl_verify_
7181 message for both "message" and "log_message" expansions, so it can be
7182 passed through if needed.
7184 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7185 new expansion $bh_ to give the encoded byte string without charset
7186 translation. Translation happens only if iconv() is available; HAVE_ICONV
7187 indicates this at build time. HEADERS_CHARSET gives the charset to
7188 translate to; headers_charset can change it in the configuration, and
7189 "headers charset" can change it in an individual filter file.
7191 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7192 that creates RFC 2047 encoded "words" labels them as that charset instead
7193 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7194 expansion operator; (ii) when Exim creates a From: line for a local
7195 message; (iii) when a header line is rewritten to include a "phrase" part.
7197 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7198 buggy, causing it to skip the first lines of messages whose message ID
7199 ended in 'D'. This would not have bitten before Exim release 4.14, because
7200 message IDs were unlikely to end in 'D' before then. The effect was to have
7201 incorrect size information for certain domains.
7203 45. #include "config.h" was missing at the start of the crypt16.c module. This
7204 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7207 46. If there was a timeout during a "random" callout check, Exim treated it as
7208 a failure of the random address, and carried on sending RSET and the real
7209 address. If the delay was just some slowness somewhere, the response to the
7210 original RCPT would be taken as a response to RSET and so on, causing
7211 mayhem of various kinds.
7213 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7214 when I implemented it. It didn't allow for the fact that some option values
7215 may legitimately be negative (e.g. size_addition), and it didn't even do
7216 the right test for positive values.
7218 48. Domain names in DNS records are case-independent. Exim always looks them up
7219 in lower case. Some resolvers return domain names in exactly the case they
7220 appear in the zone file, that is, they may contain uppercase letters. Not
7221 all resolvers do this - some return always lower case. Exim was treating a
7222 change of case by a resolver as a change of domain, similar to a widening
7223 of a domain abbreviation. This triggered its re-routing code and so it was
7224 trying to route what was effectively the same domain again. This normally
7225 caused routing to fail (because the router wouldn't handle the domain
7226 twice). Now Exim checks for this case specially, and just changes the
7227 casing of the domain that it ultimately uses when it transmits the message
7230 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7233 50. If a filter generated a file delivery with a non-absolute name (possible if
7234 no home directory exists for the router), the forbid_file option was not
7237 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7238 the existing equality matching.
7240 52. Exim was using ints instead of ino_t variables in some places where it was
7241 dealing with inode numbers.
7243 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7244 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7245 called TMPDIR, and if it finds it is different, it changes its value.
7247 54. The smtp_printf() function is now made available to local_scan() so
7248 additional output lines can be written before returning. There is also an
7249 smtp_fflush() function to enable the detection of a dropped connection.
7250 The variables smtp_input and smtp_batched_input are exported to
7253 55. Changed the default runtime configuration: the message "Unknown user"
7254 has been removed from the ACL, and instead placed on the localuser router,
7255 using the cannot_route_message feature. This means that any verification
7256 failures that generate their own messages won't get overridden. Similarly,
7257 the "Unrouteable address" message that was in the ACL for unverifiable
7258 relay addresses has also been removed.
7260 56. Added hosts_avoid_esmtp to the smtp transport.
7262 57. The exicyclog script was not checking for the esoteric option
7263 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7264 will work only if exicyclog is run under the appropriate euid.
7266 58. Following a discussion on the list, the rules by which Exim recognises line
7267 endings on incoming messages have been changed. The -dropcr and drop_cr
7268 options are now no-ops, retained only for backwards compatibility. The
7269 following line terminators are recognized: LF CRLF CR. However, special
7270 processing applies to CR:
7272 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7273 nor a local message in the state where . is a terminator.
7275 (ii) If a bare CR is encountered in a header line, an extra space is added
7276 after the line terminator so as not to end the header. The reasoning
7277 behind this is that bare CRs in header lines are most likely either
7278 to be mistakes, or people trying to play silly games.
7280 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7281 was being incorrectly given as 18 bytes larger than it should have been.
7282 This is a VOB (very old bug).
7284 60. This may never have affected anything current, but just in case it has:
7285 When the local host is found other than at the start of a list of hosts,
7286 the local host, those with the same MX, and any that follow, are discarded.
7287 When the list in question was part of a longer list of hosts, the following
7288 hosts (not currently being processed) were also being discarded. This no
7289 longer happens. I'm not sure if this situation could ever has previously
7292 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7295 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7296 header. This is something that is recommended in a new Internet Draft, and
7297 is something that is documented as being done by Sendmail. There are two
7298 possible values. For messages generated by the autoreply transport, Exim
7301 Auto-Submitted: auto-replied
7303 whereas for all other generated messages (e.g. bounces) it adds
7305 Auto-Submitted: auto-generated
7307 63. The "personal" condition in filters now includes a test for the
7308 Auto-Submitted: header. If it contains the string "auto-" the message it
7309 not considered personal.
7311 64. Added rcpt_include_affixes as a generic transport option.
7313 65. Added queue_only_override (default true).
7315 66. Added the syslog_duplication option.
7317 67. If what should have been the first header line of a message consisted of
7318 a space followed by a colon, Exim was mis-interpreting it as a header line.
7319 It isn't of course - it is syntactically invalid and should therefore be
7320 treated as the start of the message body. The misbehaviour could have
7321 caused a number of strange effects, including loss of data in subsequent
7322 header lines, and spool format errors.
7324 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7325 client host had authenticated. This control can now be exercised by an ACL
7326 for more flexibility.
7328 69. By default, callouts do not happen when testing with -bh. There is now a
7329 variant, -bhc, which does actually run the callout code, including
7330 consulting and updating the callout cache.
7332 70. Added support for saslauthd authentication, courtesy of Alexander
7335 71. If statvfs() failed on the spool or log directories while checking their
7336 size for availability, Exim confusingly gave the error "space shortage".
7337 Furthermore, in debugging mode it crashed with a floating point exception.
7338 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7339 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7340 serious problem, Exim now writes to the main and panic logs when this
7341 happens, with details of the failure. It then refuses to accept the
7342 incoming message, giving the message "spool directory problem" or "log
7343 directory problem" with a 421 code for SMTP messages.
7345 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7346 0, 1, and 2 exist, because some OS complain for execs without them (see
7347 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7348 descriptors. However, the code omitted to check that the open succeeded,
7349 causing mysterious errors if for some reason the permissions on /dev/null
7350 got screwed. Now Exim writes a message to the main and panic logs, and
7351 bombs out if it can't open /dev/null.
7353 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7354 interact so that it is all more flexible. It is supposed to remain
7355 backwards compatible. Also added extra_local_interfaces.
7357 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7358 to bomb out with an assertion failure - to the client this appears as a
7359 connection drop. This problem occurs in the part of the code that was taken
7360 from the Samba project. Fortunately, the assertion is in a very simple
7361 function, so I have fixed this by reproducing the function inline in the
7362 one place where it is called, and arranging for authentication to fail
7363 instead of killing the process with assert().
7365 75. The SPA client code was not working when the server requested OEM rather
7366 than Unicode encoding.
7368 76. Added code to make require_files with a specific uid setting more usable in
7369 the case where statting the file as root fails - usually a non-root-mounted
7370 NFS file system. When this happens and the failure is EACCES, Exim now
7371 forks a subprocess and does the per-uid checking as the relevant uid.
7373 77. Added process_log_path.
7375 78. If log_file_path was not explicitly set, a setting of check_log_space or
7376 check_log_inodes was ignored.
7378 79. If a space check for the spool or log partitions fails, the incident is now
7379 logged. Of course, in the latter case the data may get lost...
7381 80. Added the %p formatting code to string_format() so that it can be used to
7382 print addresses in debug_print(). Adjusted all the address printing in the
7383 debugging in store.c to use %p rather than %d.
7385 81. There was a concern that a line of code in smtp_in.c could overflow a
7386 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7387 initially expressed, the concern was not well-founded, because trailing
7388 spaces are removed early. However, if the trailing spaces were followed by
7389 a NULL, they did not get removed, so the overflow was possible. Two fixes
7392 (a) I re-wrote the offending code in a cleaner fashion.
7393 (b) If an incoming SMTP command contains a NULL character, it is rejected
7396 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7397 runs initgroups(), so that if the Exim user is in any additional groups,
7398 they will be used during message reception.
7404 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob