1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
59 JH/01 Remove code calling the customisable local_scan function, unless a new
60 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
62 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
63 non-signal-safe functions being used.
65 JH/03 Bug 2269: When presented with a received message having a stupidly large
66 number of DKIM-Signature headers, disable DKIM verification to avoid
67 a resource-consumption attack. The limit is set at twenty.
69 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
70 report of oldest_pass in ${authres } in consequence, and separate out
71 some descriptions of reasons for verification fail.
73 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
74 files in the spool were present and unlocked. A queue-runner could spot
75 them, resulting in a duplicate delivery. Fix that by doing the unlock
76 after the unlink. Investigation by Tim Stewart. Take the opportunity to
77 add more error-checking on spoolfile handling while that code is being
80 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
81 No known attacks, no CVE, this is defensive hardening.
83 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
84 a queue-runner could start a delivery while other operations were ongoing.
85 Cutthrough delivery was a common victim, resulting in duplicate delivery.
86 Found and investigated by Tim Stewart. Fix by using the open message data
87 file handle rather than opening another, and not locally closing it (which
88 releases a lock) for that case, while creating the temporary .eml format
89 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
91 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
92 $sender_verify_failure/$recipient_verify_failure to "random".
94 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
97 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
98 Previously this would segfault.
100 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
103 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
104 like zero, since the resolver should be doing this for us, But we need one
105 as a CNAME but no MX presence gets the CNAME returned; we need to check
106 that doesn't point to an MX to declare it "no MX returned" rather than
107 "error, loop". A new main option is added so the older capability of
108 following some limited number of chain links is maintained.
110 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
112 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
113 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
114 now-deprecated earlier definitions used only specified the range up to TLS
115 1.2 (in the older-version library docs).
117 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
119 JH/15 Rework TLS client-side context management. Stop using a global, and
120 explicitly pass a context around. This enables future use of TLS for
121 connections to service-daemons (eg. malware scanning) while a client smtp
122 connection is using TLS; with cutthrough connections this is quite likely.
124 JH/16 Fix ARC verification to do AS checks in reverse order.
126 JH/17 Support a "tls" option on the ${readsocket } expansion item.
128 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
129 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
130 Previously the "utf8" would be re-prepended for every additional message.
132 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
133 Previously thery were accepted, resulting in issues when attempting to
134 forward messages to a non-supporting MTA.
136 PP/02 Let -n work with printing macros too, not just options.
138 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
139 one parent address was copied, and bogus data was used at delivery-logging
140 time. Either a crash (after delivery) or bogus log data could result.
141 Discovery and analysis by Tim Stewart.
143 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
144 Previously if the string ended mid-character, we did not insert the
145 promised '?' replacement.
147 PP/04 Documentation: current string operators work on bytes, not codepoints.
149 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
150 should pack well giving a smaller memory footprint so better caching and
151 therefore performance. Group the declarations where this can't be done so
152 that the byte-sized flag variables are not interspersed among pointer
153 variables, giving a better chance of good packing by the compiler.
155 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
156 non-null, to avoid issues with sites running BATV. Previously reports were
157 sent with an empty envelope sender so looked like bounces.
159 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
160 The ignore_error flag wasn't being returned from the filter subprocess so
161 was not set for later routers. Investigation and fix by Matthias Kurz.
163 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
164 and a msg:complete for the whole, when a message is manually removed using
165 -Mrm. Developement by Matthias Kurz, hacked on by JH.
167 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
168 a "Gnu special" function, asprintf() in the DB utility binary builds; I
169 hope that is portable enough.
171 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
172 requiring a known-CA anchor certificate; make it now rely entirely on the
173 TLSA as an anchor. Checking the name on the leaf cert against the name
174 on the A-record for the host is still done for TA (but not for EE mode).
176 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
177 included in delivery lines for non-proxied connections, when compiled with
178 SUPPORT_SOCKS and running with proxy logging enabled.
180 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
181 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
182 move the existing event to fire before the normal logging of message
183 failure so that custom logging is bracketed by normal logging.
185 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
186 msg:fail:internal event. Developement by Matthias Kurz.
188 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
189 far too small for todays use of crypto signatures stored there. Go all
190 the way to the max DNS message size of 64kB, even though this might be
191 overmuch for IOT constrained device use.
193 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
194 copy a string over itself. The library routine is documented as not
195 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
197 JH/32 For main options check_spool_space and check_inode_space, where the
198 platform supports 64b integers, support more than the previous 2^31 kB
199 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
200 the previous G, M, k.
202 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
203 $authenticated_fail_id variable on authentication failure. Previously
206 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
207 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
208 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
209 GNUTLS_SEC_PARAM_MEDIUM.
211 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
212 side. Previously we would continue as if no SNI had been received.
214 JH/36 Harden the handling of string-lists. When a list consisted of a sole
215 "<" character, which should be a list-separator specification, we walked
216 off past the nul-terimation.
218 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
219 causes) even when the retry time is not yet met. Previously they were
220 not, meaning that when (say) an account was over-quota and temp-rejecting,
221 and multiple senders' messages were queued, only one sender would get
222 notified on each configured delay_warning cycle.
224 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
226 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
227 cipher-suites, an error can be left on the stack even for a succeeding
228 accept; this results in impossible error messages when a later operation
231 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
232 return error codes indicating retry. Under TLS1.3 this becomes required.
234 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
235 it only wrote the new authenticators, resulting in a lack of tracking of
236 peer changes of ESMTP extensions until the next cache flush.
238 JH/41 Fix the loop reading a message header line to check for integer overflow,
239 and more-often against header_maxsize. Previously a crafted message could
240 induce a crash of the recive process; now the message is cleanly rejected.
242 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
243 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
249 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
250 When redis_servers is set to a list of > 1 element, and the Redis servers
251 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
252 case of MOVED into a DEFER case instead, thus moving the query onto the
253 next server in the list. For a cluster of N elements, all N servers must
254 be defined in redis_servers.
256 GF/02 Catch and remove uninitialized value warning in exiqsumm
257 Check for existence of @ARGV before looking at $ARGV[0]
259 JH/01 Replace the store_release() internal interface with store_newblock(),
260 which internalises the check required to safely use the old one, plus
261 the allocate and data copy operations duplicated in both (!) of the
262 extant use locations.
264 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
265 modifier. This matches the restriction on the commandline.
267 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
268 Previously only the last row was returned.
270 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
271 we assumed that tags in the header were well-formed, and parsed the
272 element content after inspecting only the first char of the tag.
273 Assumptions at that stage could crash the receive process on malformed
276 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
277 While running the DKIM ACL we operate on the Permanent memory pool so that
278 variables created with "set" persist to the DATA ACL. Also (at any time)
279 DNS lookups that fail create cache records using the Permanent pool. But
280 expansions release any allocations made on the current pool - so a dnsdb
281 lookup expansion done in the DKIM ACL releases the memory used for the
282 DNS negative-cache, and bad things result. Solution is to switch to the
283 Main pool for expansions.
284 While we're in that code, add checks on the DNS cache during store_reset,
285 active in the testsuite.
286 Problem spotted, and debugging aided, by Wolfgang Breyha.
288 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
289 When none of the hosts presented to a transport match an already-open
290 connection, close it and proceed with the list. Previously we would
291 queue the message. Spotted by Lena with Yahoo, probably involving
294 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
295 Previously a spurious "250 OK id=" response was appended to the proper
298 JH/08 The "support for" informational output now, which built with Content
299 Scanning support, has a line for the malware scanner interfaces compiled
300 in. Interface can be individually included or not at build time.
302 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
303 by the template makefile "src/EDITME". The "STREAM" support for an older
304 ClamAV interface method is removed.
306 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
307 rows affected is given instead).
309 JH/11 The runtime Berkeley DB library version is now additionally output by
310 "exim -d -bV". Previously only the compile-time version was shown.
312 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
313 SMTP connection. Previously, when one had more receipients than the
314 first, an abortive onward connection was made. Move to full support for
315 multiple onward connections in sequence, handling cutthrough connection
316 for all multi-message initiating connections.
318 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
319 routers. Previously, a multi-recipient message would fail to match the
320 onward-connection opened for the first recipient, and cause its closure.
322 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
323 a timeout on read on a GnuTLS initiating connection, resulting in the
324 initiating connection being dropped. This mattered most when the callout
325 was marked defer_ok. Fix to keep the two timeout-detection methods
328 JH/15 Relax results from ACL control request to enable cutthrough, in
329 unsupported situations, from error to silently (except under debug)
330 ignoring. This covers use with PRDR, frozen messages, queue-only and
333 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
335 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
336 metadata, resulting in a crash in free().
338 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
339 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
340 Broken also in d185889f4, with init system revamp.
342 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
343 Previously we abruptly closed the connection after reading a malware-
344 found indication; now we go on to read the "scan ok" response line,
347 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
348 ACL. Previously, a crash would result.
350 JH/19 Speed up macro lookups during configuration file read, by skipping non-
351 macro text after a replacement (previously it was only once per line) and
352 by skipping builtin macros when searching for an uppercase lead character.
354 JH/20 DANE support moved from Experimental to mainline. The Makefile control
355 for the build is renamed.
357 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
358 was allocated for every new TLS startup, meaning one per message. Fix
359 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
361 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
362 reported the original. Fix to report (as far as possible) the ACL
363 result replacing the original.
365 JH/23 Fix memory leak during multi-message connections using STARTTLS under
366 OpenSSL. Certificate information is loaded for every new TLS startup,
367 and the resources needed to be freed.
369 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
371 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
374 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
375 DATA response info to the (existing) per-recipient response info for
376 the "C=" log element. It can have useful tracking info from the
377 destination system. Patch from Simon Arlott.
379 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
380 length value. Previously this would segfault.
382 HS/02 Support Avast multiline protoocol, this allows passing flags to
383 newer versions of the scanner.
385 JH/28 Ensure that variables possibly set during message acceptance are marked
386 dead before release of memory in the daemon loop. This stops complaints
387 about them when the debug_store option is enabled. Discovered specifically
388 for sender_rate_period, but applies to a whole set of variables.
389 Do the same for the queue-runner and queue-list loops, for variables set
390 from spool message files. Do the same for the SMTP per-message loop, for
391 certain variables indirectly set in ACL operations.
393 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
394 as a multi-recipient message from a mailinglist manager). The coding had
395 an arbitrary cutoff number of characters while checking for more input;
396 enforced by writing a NUL into the buffer. This corrupted long / fast
397 input. The problem was exposed more widely when more pipelineing of SMTP
398 responses was introduced, and one Exim system was feeding another.
399 The symptom is log complaints of SMTP syntax error (NUL chars) on the
400 receiving system, and refused recipients seen by the sending system
401 (propating to people being dropped from mailing lists).
402 Discovered and pinpointed by David Carter.
404 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
405 replaced by the ${authresults } expansion.
407 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
409 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
410 allows proper process termination in container environments.
412 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
413 Previously the "final dot" had a newline after it; ensure it is CR,LF.
415 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
416 and "err_perm", deprecated since 4.83 when the RFC-defined words
417 "temperror" and "permerror" were introduced.
419 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
420 transport-filters or DKIM-signing. The restriction was lost in the
421 consolidation of verify-callout and delivery SMTP handling.
422 Extend the restriction to also cover ARC-signing.
424 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
425 in defer=pass mode supply a 450 to the initiator. Previously the message
428 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
429 tls_require_ciphers is used as before.
431 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
432 "pass_unscanned". Only tmpfails from the scanner are written to
433 the paniclog, as they may require admin intervention (permission
434 denied, license issues). Other scanner errors (like decompression
435 bombs) do not cause a paniclog entry.
437 JH/36 Fix reinitialisation of DKIM logging variable between messages.
438 Previously it was possible to log spurious information in receive log
441 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
442 triggered odd behaviour from Outlook Express clients.
444 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
447 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
448 since the IETF WG has not yet settled on that versus the original
449 "bare" representation.
451 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
452 Previously the millisecond value corrupted the output.
453 Fix also for syslog_pid=no and log_selector +pid, for which the pid
454 corrupted the output.
460 JH/01 Rework error string handling in TLS interface so that the caller in
461 more cases is responsible for logging. This permits library-sourced
462 string to be attached to addresses during delivery, and collapses
463 pairs of long lines into single ones.
465 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
466 during configuration. Wildcards are allowed and expanded.
468 JH/02 Rework error string handling in DKIM to pass more info back to callers.
469 This permits better logging.
471 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
472 do not close it down and have the child transport start it up again on
473 the passed-on TCP connection. Instead, proxy the child (and any
474 subsequent ones) for TLS via a unix-domain socket channel. Logging is
475 affected: the continued delivery log lines do not have any DNSSEC, TLS
476 Certificate or OCSP information. TLS cipher information is still logged.
478 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
479 identical IP addresses on different listening ports. Will also affect
482 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
483 add noisy ifdef guards to special-case this sillyness.
484 Patch from Bernd Kuhls.
486 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
487 than 255 are no longer allowed.
489 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
490 Disable the session-cache too, which might reduce our load. Since we
491 currrectly use a new context for every connection, both as server and
492 client, there is no benefit for these.
493 GnuTLS appears to not support tickets server-side by default (we don't
494 call gnutls_session_ticket_enable_server()) but client side is enabled
495 by default on recent versions (3.1.3 +) unless the PFS priority string
498 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
499 <https://reproducible-builds.org/specs/source-date-epoch/>.
501 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
502 the check for any unsuccessful recipients did not notice the limit, and
503 erroneously found still-pending ones.
505 JH/08 Pipeline CHUNKING command and data together, on kernels that support
506 MSG_MORE. Only in-clear (not on TLS connections).
508 JH/09 Avoid using a temporary file during transport using dkim. Unless a
509 transport-filter is involved we can buffer the headers in memory for
510 creating the signature, and read the spool data file once for the
511 signature and again for transmission.
513 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
514 4.77 as the kernel support then wasn't solid, having issues in 64bit
515 mode. Now, it's been long enough. Add support for FreeBSD also.
517 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
518 case where the routing stage had gathered several addresses to send to
519 a host before calling the transport for the first, we previously failed
520 to close down TLS in the old transport process before passing the TCP
521 connection to the new process. The new one sent a STARTTLS command
522 which naturally failed, giving a failed delivery and bloating the retry
523 database. Investigation and fix prototype from Wolfgang Breyha.
525 JH/12 Fix check on SMTP command input synchronisation. Previously there were
526 false-negatives in the check that the sender had not preempted a response
527 or prompt from Exim (running as a server), due to that code's lack of
528 awareness of the SMTP input buffering.
530 PP/04 Add commandline_checks_require_admin option.
531 Exim drops privileges sanely, various checks such as -be aren't a
532 security problem, as long as you trust local users with access to their
533 own account. When invoked by services which pass untrusted data to
534 Exim, this might be an issue. Set this option in main configuration
535 AND make fixes to the calling application, such as using `--` to stop
538 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
539 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
540 responses to those, into a single TLS record each way (this usually means
541 a single packet). As a side issue, smtp_enforce_sync now works on TLS
544 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
545 affects you only if you're dancing at the edge of the param size limits.
546 If you are, and this message makes sense to you, then: raise the
547 configured limit or use OpenSSL 1.1. Nothing we can do for older
550 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
551 cmdline element to get the documented default one. Previously it was
554 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
555 are unsuable for non-message-specific reasons - eg. connection timeout,
556 banner-time rejection.
558 JH/16 Fix logging of delivery remote port, when specified by router, under
561 PP/06 Repair manualroute's ability to take options in any order, even if one
562 is the name of a transport.
565 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
567 JH/17 Change the list-building routines interface to use the expanding-string
568 triplet model, for better allocation and copying behaviour.
570 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
571 Previously it was constructed the first time a possibly-matching string
572 was met in the configuration file input during startup; now it is done
575 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
576 compatible one, to avoid the (poorly documented) possibility of a config
577 file in the working directory redirecting the DB files, possibly correpting
578 some existing file. CVE-2017-10140 assigned for BDB.
580 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
581 cache-hot. Previously, although the result was properly cached, the
582 initial verify call returned a defer.
584 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
585 the main verify for receipient in uncached-mode.
587 JH/22 Retire historical build files to an "unsupported" subdir. These are
588 defined as "ones for which we have no current evidence of testing".
590 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
591 if present. Previously it was ignored.
593 JH/24 Start using specified-initialisers in C structure init coding. This is
594 a C99 feature (it's 2017, so now considered safe).
596 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
597 if was a fixed-sized field and bitmask ops via macros; it is now more
600 PP/07 GitHub PR 56: Apply MariaDB build fix.
601 Patch provided by Jaroslav Škarvada.
603 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
604 during Coverity cleanups [4.87 JH/47]
605 Diagnosis and fix provided by Michael Fischer v. Mollard.
607 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
608 the right size to place the terminating semicolon on its own folded
609 line, the header hash was calculated to an incorrect value thanks to
610 the (relaxed) space the fold became.
612 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
613 and confused the parent.
615 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
616 which could crash as a result. This could lead to undeliverable messages.
618 JH/28 Logging: "next input sent too soon" now shows where input was truncated
621 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
622 matters on fast-turnover and PID-randomising systems, which were getting
623 out-of-order delivery.
625 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
626 a possibly-overlapping copy. The symptom was that "Remote host closed
627 connection in response to HELO" was logged instead of the actual 4xx
630 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
631 Previously only that bufferd was discarded, resulting in SYMTP command
634 JH/32 DKIM: when a message has multiple signatures matching an identity given
635 in dkim_verify_signers, run the dkim acl once for each. Previously only
636 one run was done. Bug 2189.
638 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
639 "panic the current process" to "deliberately defer". The panic log is
640 still written with the problem list name; the mail and reject logs now
641 get a temp-reject line for the message that was being handled, saying
642 something like "domains check lookup or other defer". The SMTP 451
643 message is still "Temporary local problem".
645 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
646 A crafted sequence of BDAT commands could result in in-use memory beeing
647 freed. CVE-2017-16943.
649 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
650 from SMTP input. Previously it was always done; now only done for DATA
651 and not BDAT commands. CVE-2017-16944.
653 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
654 to the message (such as an overlong header line). Previously this was
655 not done and we did not exit BDAT mode. Followon from the previous item
656 though a different problem.
662 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
663 than -2003 did; needs libidn2 in addition to libidn.
665 JH/02 The path option on a pipe transport is now expanded before use.
667 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
668 Patch provided by "Björn", documentation fix added too.
670 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
671 missing a wire-to-host endian conversion.
673 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
674 close after a BDAT command line could be taken as a following command,
675 giving a synch failure. Fix by only checking for synch immediately
676 before acknowledging the chunk.
678 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
679 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
680 macro. Patches provided by Josh Soref.
682 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
683 Previously we did not; the RFC seems ambiguous and VRFY is not listed
684 by IANA as a service extension. However, John Klensin suggests that we
687 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
688 the dkim code may be unix-mode line endings rather than smtp wire-format
689 CRLF, so prepend a CR to any bare LF.
691 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
692 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
694 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
695 the first were themselves being wrongly included in the feed into dkim
696 processing; with most chunk sizes in use this resulted in an incorrect
697 body hash calculated value.
699 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
700 DKIM signature block, for verification. Although advised against by
701 standards it is specifically not ruled illegal.
703 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
705 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
706 missing a body hash (the bh= tag).
708 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
709 It seems that HAProxy sends the Proxy Protocol information in clear and
710 only then does a TLS startup, so do the same.
712 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
713 TCP connections (such as for Spamd) unless the daemon successfully set
714 Fast Open mode on its listening sockets. This fixes breakage seen on
715 too-old kernels or those not configured for Fast Open, at the cost of
716 requiring both directions being enabled for TFO, and TFO never being used
717 by non-daemon-related Exim processes.
719 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
720 endings, at least on the first header line. Try to canonify any that get
721 past that check, despite the cost.
723 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
724 now limited to an arbitrary five deep, while parsing addresses with the
725 strip_excess_angle_brackets option enabled.
727 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
728 instead leave the unprompted TLS handshake in socket buffer for the
729 TLS library to consume.
731 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
733 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
735 JH/16 Drop variables when they go out of scope. Memory management drops a whole
736 region in one operation, for speed, and this leaves assigned pointers
737 dangling. Add checks run only under the testsuite which checks all
738 variables at a store-reset and panics on a dangling pointer; add code
739 explicitly nulling out all the variables discovered. Fixes one known
740 bug: a transport crash, where a dangling pointer for $sending_ip_address
741 originally assigned in a verify callout, is re-used.
743 PP/06 Drop '.' from @INC in various Perl scripts.
745 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
747 PP/08 Reduce a number of compilation warnings under clang; building with
748 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
749 should be warning-free.
751 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
753 HS/01 Fix portability problems introduced by PP/08 for platforms where
754 realloc(NULL) is not equivalent to malloc() [SunOS et al].
756 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
757 chunk. This allows us to accept broken chunked messages. We need a more
758 general solution here.
760 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
761 already-broken messages in the queue.
763 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
765 JH/19 Fix reference counting bug in routing-generated-address tracking.
771 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
772 supports it and a size is available (ie. the sending peer gave us one).
774 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
775 years of being deprecated). The replacements are the ACLs
776 acl_smtp_mime and acl_not_smtp_mime.
778 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
779 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
780 or even an in-clear connection were permitted. Now, if the host lookup
781 was dnssec and dane was requested then the host is only used if the
782 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
783 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
784 if one fails this test.
785 This means that a poorly-configured remote DNS will make it incommunicado;
786 but it protects against a DNS-interception attack on it.
788 JH/04 Bug 1810: make continued-use of an open smtp transport connection
789 non-noisy when a race steals the message being considered.
791 JH/05 If main configuration option tls_certificate is unset, generate a
792 self-signed certificate for inbound TLS connections.
794 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
795 in rewrites and routers.
797 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
798 and logged a warning sing 4.83; now they are a configuration file error.
800 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
801 (lacking @domain). Apply the same qualification processing as RCPT.
803 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
805 JH/10 Support ${sha256:} applied to a string (as well as the previous
808 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
809 a cutthrough deliver is pending, as we always want to make a connection.
810 This also avoids re-routing the message when later placing the cutthrough
811 connection after a verify cache hit.
812 Do not update it with the verify result either.
814 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
815 when routing results in more than one destination address.
817 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
818 signing (which inhibits the cutthrough capability). Previously only
819 the presence of an option was tested; now an expansion evaluating as
820 empty is permissible (obviously it should depend only on data available
821 when the cutthrough connection is made).
823 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
824 the relevant preceding SMTP command did not note the pipelining mode.
826 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
827 Previously they were not counted.
829 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
830 as one having no matching records. Previously we deferred the message
831 that needed the lookup.
833 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
834 distinguished as "(=".
836 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
837 for missing MX records. Previously it only worked for missing A records.
839 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
841 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
842 after the data-go-ahead and data-ack. Patch from Jason Betts.
844 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
845 even for a "none" policy. Patch from Tony Meyer.
847 JH/22 Fix continued use of a connection for further deliveries. If a port was
848 specified by a router, it must also match for the delivery to be
851 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
852 When one of the recipients of a message was unsuitable for the connection
853 (has no matching addresses), we lost track of needing to mark it
854 deferred. As a result mail would be lost.
856 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
858 JH/25 Decoding ACL controls is now done using a binary search; the source code
859 takes up less space and should be simpler to maintain. Merge the ACL
860 condition decode tables also, with similar effect.
862 JH/26 Fix problem with one_time used on a redirect router which returned the
863 parent address unchanged. A retry would see the parent address marked as
864 delivered, so not attempt the (identical) child. As a result mail would
867 JH/27 Fix a possible security hole, wherein a process operating with the Exim
868 UID can gain a root shell. Credit to http://www.halfdog.net/ for
869 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
872 JH/28 Enable {spool,log} filesystem space and inode checks as default.
873 Main config options check_{log,spool}_{inodes,space} are now
874 100 inodes, 10MB unless set otherwise in the configuration.
876 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
877 Previously it only applied to the main-section connection policy
880 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
882 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
883 by me. Added RFC7919 DH primes as an alternative.
885 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
886 are not in the system include path.
888 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
889 GnuTLS, when a session startup failed (eg because the client disconnected)
890 Exim did stdio operations after fclose. This was exposed by a recent
891 change which nulled out the file handle after the fclose.
893 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
894 signed directly by the cert-signing cert, rather than an intermediate
895 OCSP-signing cert. This is the model used by LetsEncrypt.
897 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
899 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
900 an incoming connection.
902 HS/02 Bug 1802: Do not half-close the connection after sending a request
905 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
906 fallback to "prime256v1".
908 JH/34 SECURITY: Use proper copy of DATA command in error message.
909 Could leak key material. Remotely exploitable. CVE-2016-9963.
915 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
916 and 3.4.4 - once the server is enabled to respond to an OCSP request
917 it does even when not requested, resulting in a stapling non-aware
918 client dropping the TLS connection.
920 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
921 support variable-length bit vectors. No functional change.
923 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
924 The I= interface field on outgoing lines is now after the H= remote
925 host field, same as incoming lines. There is a separate
926 outgoing_interface log selector which allows you to disable the
929 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
930 If not running log_selector +smtp_connection the mainlog would be held
931 open indefinitely after a "too many connections" event, including to a
932 deleted file after a log rotate. Leave the per net connection logging
933 leaving it open for efficiency as that will be quickly detected by the
934 check on the next write.
936 HS/01 Bug 1671: Fix post transport crash.
937 Processing the wait-<transport> messages could crash the delivery
938 process if the message IDs didn't exist for some reason. When
939 using 'split_spool_directory=yes' the construction of the spool
940 file name failed already, exposing the same netto behaviour.
942 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
943 mime_regex ACL conditions.
945 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
946 to DSN fail messages (bounces): remote IP, remote greeting, remote response
947 to HELO, local diagnostic string.
949 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
950 log line to debug. Even when configured with a tls authenticator many
951 client connections are expected to not authenticate in this way, so
952 an authenticate fail is not an error.
954 HS/02 Add the Exim version string to the process info. This way exiwhat
955 gives some more detail about the running daemon.
957 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
958 matter for fast-change records such as DNSBLs.
960 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
961 retry record, even if constant. There may be multiple transports with
962 different interface settings and the retry behaviour needs to be kept
965 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
967 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
969 JH/10 Bug 840: fix log_defer_output option of pipe transport
971 JH/11 Bug 830: use same host for all RCPTS of a message, even under
972 hosts_randomize. This matters a lot when combined with mua_wrapper.
974 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
975 ${quote_pgsql:<string>} operator.
977 JH/13 Bug 1708: avoid misaligned access in cached lookup.
979 JH/14 Change header file name for freeradius-client. Relevant if compiling
980 with Radius support; from the Gentoo tree and checked under Fedora.
982 JH/15 Bug 1712: Introduce $prdr_requested flag variable
984 JH/16 Bug 1714: Permit an empty string as expansion result for transport
985 option transport_filter, meaning no filtering.
987 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
989 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
990 defaults to "*" (all hosts). The variable is now available when not built
991 with TLS, default unset, mainly to enable keeping the testsuite sane.
992 If a server certificate is not supplied (via tls_certificate) an error is
993 logged, and clients will find TLS connections fail on startup. Presumably
994 they will retry in-clear.
995 Packagers of Exim are strongly encouraged to create a server certificate
996 at installation time.
998 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
999 with the $config_file variable.
1001 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1002 in transport context, after the attempt, and per-recipient. The latter type
1003 is per host attempted. The event data is the error message, and the errno
1004 information encodes the lookup type (A vs. MX) used for the (first) host,
1005 and the trailing two digits of the smtp 4xx response.
1007 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1008 to write to mainlog (or rejectlog, paniclog) in the window between file
1009 creation and permissions/ownership being changed. Particularly affects
1010 installations where exicyclog is run as root, rather than exim user;
1011 result is that the running daemon panics and dies.
1013 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1015 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1016 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1017 "pri" and "weight". Note that the previous implicit priority given by the
1018 list order is no longer honoured.
1020 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1021 for DKIM processing.
1023 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1024 by defining SUPPORT_SOCKS.
1026 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1027 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1028 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1029 variables are renamed to proxy_{local,external}_{address,port}.
1031 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1032 for a build by defining SUPPORT_I18N
1034 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1035 of the query string, and make ${quote_redis:} do that quoting.
1037 JH/29 Move Events support from Experimental to mainline, enabled by default
1038 and removable for a build by defining DISABLE_EVENT.
1040 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1042 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1043 cached by the daemon.
1045 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1046 by defining LOOKUP_REDIS. The libhiredis library is required.
1048 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1049 keys are given for lookup.
1051 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1052 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1053 only supported when built with TLS support. The PolarSSL SHA routines
1054 are still used when the TLS library is too old for convenient support.
1056 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1057 openssl_options), for security. OpenSSL forces this from version 1.1.0
1058 server-side so match that on older versions.
1060 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1061 allocation for $value could be released as the expansion processing
1062 concluded, but leaving the global pointer active for it.
1064 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1065 and to use the domains and local_parts ACL conditions.
1067 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1068 incorrectly not doubled on cutthrough transmission, hence seen as a
1069 body-termination at the receiving system - resulting in truncated mails.
1070 Commonly the sender saw a TCP-level error, and retransmitted the message
1071 via the normal store-and-forward channel. This could result in duplicates
1072 received - but deduplicating mailstores were liable to retain only the
1073 initial truncated version.
1075 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1077 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1079 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1080 we're in there, support oversigning also; bug 1309.
1082 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1084 HS/04 Add support for keep_environment and add_environment options.
1086 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1087 either intentional arithmetic overflow during PRNG, or testing config-
1090 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1091 delivery resulted in actual delivery. Cancel cutthrough before DATA
1094 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1095 rejects a recipient: pass the reject to the originator.
1097 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1098 Many were false-positives and ignorable, but it's worth fixing the
1101 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1102 for the new environment-manipulation done at startup. Move the routines
1103 from being local to tls.c to being global via the os.c file.
1105 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1106 an extract embedded as result-arg for a map, the first arg for extract
1107 is unavailable so we cannot tell if this is a numbered or keyed
1108 extraction. Accept either.
1114 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1117 JH/02 The smtp transport option "multi_domain" is now expanded.
1119 JH/03 The smtp transport now requests PRDR by default, if the server offers
1122 JH/04 Certificate name checking on server certificates, when exim is a client,
1123 is now done by default. The transport option tls_verify_cert_hostnames
1124 can be used to disable this per-host. The build option
1125 EXPERIMENTAL_CERTNAMES is withdrawn.
1127 JH/05 The value of the tls_verify_certificates smtp transport and main options
1128 default to the word "system" to access the system default CA bundle.
1129 For GnuTLS, only version 3.0.20 or later.
1131 JH/06 Verification of the server certificate for a TLS connection is now tried
1132 (but not required) by default. The verification status is now logged by
1133 default, for both outbound TLS and client-certificate supplying inbound
1136 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1139 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1140 Status Notification (bounce) messages are now MIME format per RFC 3464.
1141 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1142 under the control of the dsn_advertise_hosts option, and routers may
1143 have a dsn_lasthop option.
1145 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1146 default, modifiable by a malware= option. The list separator for
1147 the options can now be changed in the usual way. Bug 68.
1149 JH/10 The smtp_receive_timeout main option is now expanded before use.
1151 JH/11 The incoming_interface log option now also enables logging of the
1152 local interface on delivery outgoing connections.
1154 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1155 if the interface and destination host and port all match.
1157 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1160 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1161 Patch from Andrew Lewis.
1163 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1164 now supports optional time-restrictions, weighting, and priority
1165 modifiers per server. Patch originally by <rommer@active.by>.
1167 JH/16 The spamd_address main option now supports a mixed list of local
1168 and remote servers. Remote servers can be IPv6 addresses, and
1169 specify a port-range.
1171 JH/17 Bug 68: The spamd_address main option now supports an optional
1172 timeout value per server.
1174 JH/18 Bug 1581: Router and transport options headers_add/remove can
1175 now have the list separator specified.
1177 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1180 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1183 JH/21 Support for the A6 type of dns record is withdrawn.
1185 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1186 rather than the verbs used.
1188 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1189 from 255 to 1024 chars.
1191 JH/24 Verification callouts now attempt to use TLS by default.
1193 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1194 are generic router options now. The defaults didn't change.
1196 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1197 Original patch from Alexander Shikoff, worked over by JH.
1199 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1200 files if ZCAT_COMMAND is not executable.
1202 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1204 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1206 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1207 Normally benign, it bites when the pair was led to by a CNAME;
1208 modern usage is to not canonicalize the domain to a CNAME target
1209 (and we were inconsistent anyway for A-only vs AAAA+A).
1211 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1213 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1214 when evaluating $sender_host_dnssec.
1216 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1217 $sender_helo_dnssec variable.
1219 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1221 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1223 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1225 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1226 documented as working, but never had. Support all but $spam_report.
1228 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1229 added for tls authenticator.
1231 HS/03 Add perl_taintmode main config option
1237 TL/01 When running the test suite, the README says that variables such as
1238 no_msglog_check are global and can be placed anywhere in a specific
1239 test's script, however it was observed that placement needed to be near
1240 the beginning for it to behave that way. Changed the runtest perl
1241 script to read through the entire script once to detect and set these
1242 variables, reset to the beginning of the script, and then run through
1243 the script parsing/test process like normal.
1245 TL/02 The BSD's have an arc4random API. One of the functions to induce
1246 adding randomness was arc4random_stir(), but it has been removed in
1247 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1248 function when detected.
1250 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1251 cause callback expansion.
1253 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1254 syntax errors in an expansion can be treated as a string instead of
1255 logging or causing an error, due to the internal use of bool_lax
1256 instead of bool when processing it.
1258 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1259 server certificates when making smtp deliveries.
1261 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1263 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1265 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1267 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1268 Merged patch from Sebastian Wiedenroth.
1270 JH/05 Fix results-pipe from transport process. Several recipients, combined
1271 with certificate use, exposed issues where response data items split
1272 over buffer boundaries were not parsed properly. This eventually
1273 resulted in duplicates being sent. This issue only became common enough
1274 to notice due to the introduction of connection certificate information,
1275 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1277 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1278 size buffer was used, resulting in syntax errors when an expansion
1281 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1282 version 3.3.6 or later.
1284 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1285 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1286 both become "event_action", the variables become $event_name, $event_data
1287 and $event_defer_errno. There is a new variable $verify_mode, usable in
1288 routers, transports and related events. The tls:cert event is now also
1289 raised for inbound connections, if the main configuration event_action
1292 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1293 early OCSP support, but no stapling (appears to be less than 1.0.0).
1295 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1296 server certificate names available under the smtp transport option
1297 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1300 JH/10 Time-related extraction expansions from certificates now use the main
1301 option "timezone" setting for output formatting, and are consistent
1302 between OpenSSL and GnuTLS compilations. Bug 1541.
1304 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1305 encoded parameter in the incoming message. Bug 1558.
1307 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1308 include certificate info, eximon was claiming there were spoolfile
1311 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1313 JH/14 Log delivery-related information more consistently, using the sequence
1314 "H=<name> [<ip>]" wherever possible.
1316 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1317 are problematic for Debian distribution, omit them from the release
1320 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1322 JH/16 Fix string representation of time values on 64bit time_t architectures.
1325 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1326 output list separator was used.
1331 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1332 checkers that were complaining about end of non-void function with no
1335 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1336 This was a regression introduced in 4.83 by another bugfix.
1338 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1340 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1341 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1347 TF/01 Correctly close the server side of TLS when forking for delivery.
1349 When a message was received over SMTP with TLS, Exim failed to clear up
1350 the incoming connection properly after forking off the child process to
1351 deliver the message. In some situations the subsequent outgoing
1352 delivery connection happened to have the same fd number as the incoming
1353 connection previously had. Exim would try to use TLS and fail, logging
1354 a "Bad file descriptor" error.
1356 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1357 utilities have not been installed.
1359 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1360 temporary space as the ACL may create new global variables.
1362 TL/01 LDAP support uses per connection or global context settings, depending
1363 upon the detected version of the libraries at build time.
1365 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1366 to extract and use the src ip:port in logging and expansions as if it
1367 were a direct connection from the outside internet. PPv2 support was
1368 updated based on HAProxy spec change in May 2014.
1370 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1372 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1373 Properly escape header and check for NULL return.
1375 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1378 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1380 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1381 characters in header names, implemented as a verify condition.
1382 Contributed by Michael Fischer v. Mollard.
1384 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1385 results permerror and temperror. Previous values are deprecated but
1386 still accepted. In a future release, err_perm and err_temp will be
1387 completely removed, which will be a backward incompatibility if the
1388 ACL tests for either of these two old results. Patch contributed by
1389 user bes-internal on the mailing list.
1391 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1393 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1394 selectors, in both main and reject logs.
1396 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1399 JH/07 Add malware type "sock" for talking to simple daemon.
1401 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1403 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1404 routers/transports under cutthrough routing.
1406 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1407 numbers. Touch up "bool" conditional to keep the same definition.
1409 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1411 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1413 JH/12 Expand items in router/transport headers_add or headers_remove lists
1414 individually rather than the list as a whole. Bug 1452.
1416 Required for reasonable handling of multiple headers_ options when
1417 they may be empty; requires that headers_remove items with embedded
1418 colons must have them doubled (or the list-separator changed).
1420 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1421 view the policy declared in the DMARC record. Currently, $dmarc_status
1422 is a combined value of both the record presence and the result of the
1425 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1427 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1428 dnslookup router and the smtp transport (applying to the forward
1431 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1432 of ldap servers used for a specific lookup. Patch provided by Heiko
1435 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1436 New variable $lookup_dnssec_authenticated for observability.
1438 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1439 Patch submitted by Lars Timman.
1441 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1443 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1444 Requires trusted mode and valid format message id, aborts otherwise.
1445 Patch contributed by Heiko Schlichting.
1447 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1448 certextract with support for various fields. Bug 1358.
1450 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1451 is requested by default, modifiable by smtp transport option
1454 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1455 operate on certificate variables to give certificate fingerprints
1456 Also new ${sha256:cert_variable}.
1458 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1460 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1461 Christian Aistleitner.
1463 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1465 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1466 file. Patch from Wolfgang Breyha.
1468 JH/25 Expand the coverage of the delivery $host and $host_address to
1469 client authenticators run in verify callout. Bug 1476.
1471 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1472 align with daemon_smtp_ports. Bug 72.
1474 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1475 support and error reporting did not work properly.
1477 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1478 and is readable. Patch from Andrew Colin Kissa.
1480 TL/14 Enhance documentation of ${run expansion and how it parses the
1481 commandline after expansion, particularly in the case when an
1482 unquoted variable expansion results in an empty value.
1484 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1486 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1487 by renaming away from it. Addresses GH issue 15, reported by
1490 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1492 TL/15 SECURITY: prevent double expansion in math comparison functions
1493 (can expand unsanitized data). Not remotely exploitable.
1500 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1502 PP/02 Make -n do something, by making it not do something.
1503 When combined with -bP, the name of an option is not output.
1505 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1508 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1509 $sender_host_name and config options to manage this, and basic check
1512 PP/05 DSCP support for outbound connections and control modifier for inbound.
1514 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1515 (Only plugin which currently uses this is kerberos4, which nobody should
1516 be using, but we should make it available and other future plugins might
1517 conceivably use it, even though it would break NAT; stuff *should* be
1518 using channel bindings instead).
1520 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1521 name; added for Sendmail compatibility; requires admin caller.
1522 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1523 just ignore it); requires trusted caller.
1524 Also parse but ignore: -Ac -Am -X<logfile>
1527 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1529 TL/02 Add +smtp_confirmation as a default logging option.
1531 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1532 Patch by Magnus Holmgren from 2007-02-20.
1534 TL/04 Bugzilla 1281 - Spec typo.
1535 Bugzilla 1283 - Spec typo.
1536 Bugzilla 1290 - Spec grammar fixes.
1538 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1540 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1542 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1543 mailing list by by Dmitry Isaikin.
1545 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1547 JH/02 Support "G" suffix to numbers in ${if comparisons.
1549 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1551 NM/01 Bugzilla 1197 - Spec typo
1552 Bugzilla 1196 - Spec examples corrections
1554 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1556 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1557 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1560 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1561 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1563 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1564 "acl {{name}{arg}...}", and optional args on acl condition
1567 JH/05 Permit multiple router/transport headers_add/remove lines.
1569 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1571 JH/07 Avoid using a waiting database for a single-message-only transport.
1572 Performance patch from Paul Fisher. Bugzilla 1262.
1574 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1577 JH/09 Add $headers_added variable, with content from use of ACL modifier
1578 add_header (but not yet added to the message). Bugzilla 199.
1580 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1581 Pulled from Bugzilla 817 by Wolfgang Breyha.
1583 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1585 (nb: this is the same fix as in Exim 4.80.1)
1587 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1590 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1593 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1595 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1596 advertise SMTP AUTH mechanism to us, instead of a generic
1597 protocol violation error. Also, make Exim more robust to bad
1598 data from the Dovecot auth socket.
1600 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1602 When a queue runner is handling a message, Exim first routes the
1603 recipient addresses, during which it prunes them based on the retry
1604 hints database. After that it attempts to deliver the message to
1605 any remaining recipients. It then updates the hints database using
1608 So if a recipient address works intermittently, it can get repeatedly
1609 deferred at routing time. The retry hints record remains fresh so the
1610 address never reaches the final cutoff time.
1612 This is a fairly common occurrence when a user is bumping up against
1613 their storage quota. Exim had some logic in its local delivery code
1614 to deal with this. However it did not apply to per-recipient defers
1615 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1617 This change adds a proper retry rule check during routing so that the
1618 final cutoff time is checked against the message's age. We only do
1619 this check if there is an address retry record and there is not a
1620 domain retry record; this implies that previous attempts to handle
1621 the address had the retry_use_local_parts option turned on. We use
1622 this as an approximation for the destination being like a local
1623 delivery, as in LMTP.
1625 I suspect this new check makes the old local delivery cutoff check
1626 redundant, but I have not verified this so I left the code in place.
1628 TF/02 Correct gecos expansion when From: is a prefix of the username.
1630 Test 0254 submits a message to Exim with the header
1634 When I ran the test suite under the user fanf2, Exim expanded
1635 the header to contain my full name, whereas it should have added
1636 a Resent-Sender: header. It erroneously treats any prefix of the
1637 username as equal to the username.
1639 This change corrects that bug.
1641 GF/01 DCC debug and logging tidyup
1642 Error conditions log to paniclog rather than rejectlog.
1643 Debug lines prefixed by "DCC: " to remove any ambiguity.
1645 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1647 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1648 Bug spotted by Jeremy Harris; was flawed since initial commit.
1649 Would have resulted in OCSP responses post-SNI triggering an Exim
1650 NULL dereference and crash.
1652 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1654 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1655 Bug detection, analysis and fix by Samuel Thibault.
1656 Bugzilla 1331, Debian bug #698092.
1658 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1660 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1661 Server implementation by Todd Lyons, client by JH.
1662 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1663 config variable "prdr_enable" controls whether the server
1664 advertises the facility. If the client requests PRDR a new
1665 acl_data_smtp_prdr ACL is called once for each recipient, after
1666 the body content is received and before the acl_smtp_data ACL.
1667 The client is controlled by both of: a hosts_try_prdr option
1668 on the smtp transport, and the server advertisement.
1669 Default client logging of deliveries and rejections involving
1670 PRDR are flagged with the string "PRDR".
1672 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1673 fclose(). Diagnosis by Todd Lyons.
1675 PP/17 Update configure.default to handle IPv6 localhost better.
1676 Patch by Alain Williams (plus minor tweaks).
1679 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1680 This is now consistent with GnuTLS, and is now documented: the
1681 previous undocumented portable approach to treating the option as
1682 unset was to force an expansion failure. That still works, and
1683 an empty string is now equivalent.
1685 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1686 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1687 not performing validation itself.
1689 PP/20 Added force_command boolean option to pipe transport.
1690 Patch from Nick Koston, of cPanel Inc.
1692 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1695 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1697 PP/21 Fix eximon continuous updating with timestamped log-files.
1698 Broken in a format-string cleanup in 4.80, missed when I repaired the
1699 other false fix of the same issue.
1700 Report and fix from Heiko Schlichting.
1703 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1704 Report from Prashanth Katuri.
1706 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1707 It's SecureTransport, so affects any MacOS clients which use the
1708 system-integrated TLS libraries, including email clients.
1710 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1711 using a MIME ACL for non-SMTP local injection.
1712 Report and assistance in diagnosis by Warren Baker.
1714 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1716 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1718 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1719 last id that failed so it may be referenced in subsequent ACL's.
1721 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1724 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1725 ldap library initialization, allowing self-signed CA's to be
1726 used. Also properly sets require_cert option later in code by
1727 using NULL (global ldap config) instead of ldap handle (per
1728 session). Bug diagnosis and testing by alxgomz.
1730 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1731 the src/util/ subdirectory.
1733 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1734 renamed to Transport Post Delivery Action by Jeremy Harris, as
1737 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1738 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1739 redis_servers = needs to be configured which will be used by the redis
1740 lookup. Patch from Warren Baker, of The Packet Hub.
1742 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1744 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1745 hostname or reverse DNS when processing a host list. Used suggestions
1746 from multiple comments on this bug.
1748 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1750 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1751 Missed a few lines, added it to make the runtest require no keyboard
1754 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1755 contains upper case chars. Make router use caseful_local_part.
1757 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1758 support when GnuTLS has been built with p11-kit.
1764 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1766 This, or similar/improved, will also be change PP/11 of 4.82.
1772 PP/01 Handle short writes when writing local log-files.
1773 In practice, only affects FreeBSD (8 onwards).
1774 Bugzilla 1053, with thanks to Dmitry Isaikin.
1776 NM/01 Bugzilla 949 - Documentation tweak
1778 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1781 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1783 PP/02 Implemented gsasl authenticator.
1785 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1787 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1788 `pkg-config foo` for cflags/libs.
1790 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1791 with rest of GSASL and with heimdal_gssapi.
1793 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1794 `pkg-config foo` for cflags/libs for the TLS implementation.
1796 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1797 properties get this fed in as external SSF. A number of robustness
1798 and debugging improvements to the cyrus_sasl authenticator.
1800 PP/08 cyrus_sasl server now expands the server_realm option.
1802 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1803 Patch by Jeremy Harris.
1805 PP/10 Added dbmjz lookup type.
1807 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1809 PP/12 MAIL args handles TAB as well as SP, for better interop with
1810 non-compliant senders.
1811 Analysis and variant patch by Todd Lyons.
1813 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1814 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1815 Patch from Dirk Mueller <dmueller@suse.com>
1817 PP/13 tls_peerdn now print-escaped for spool files.
1818 Observed some $tls_peerdn in wild which contained \n, which resulted
1819 in spool file corruption.
1821 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1822 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1823 or write after TLS renegotiation, which otherwise led to messages
1826 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1827 as a tracking header (ie: a signed header comes before the signature).
1828 Patch from Wolfgang Breyha.
1830 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1831 comma-sep list; embedded commas doubled.
1833 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1835 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1837 Report and patch from Dmitry Banschikov.
1839 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1840 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1841 failures. We appear to now support TLS1.1+ with Exim.
1843 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1844 lets Exim select keys and certificates based upon TLS SNI from client.
1845 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1846 before an outbound SMTP session. New log_selector, +tls_sni.
1848 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1849 NULL dereference. Report and patch from Alun Jones.
1851 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1852 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1853 Not seeing resolver debug output on NetBSD, but suspect this is a
1854 resolver implementation change.
1856 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1857 Left warnings. Added "eximon gdb" invocation mode.
1859 PP/21 Defaulting "accept_8bitmime" to true, not false.
1861 PP/22 Added -bw for inetd wait mode support.
1863 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1864 locate the relevant includes and libraries. Made this the default.
1866 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1867 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1869 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1870 This may cause build issues on older platforms.
1872 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1873 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1874 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1875 Added SNI support via GnuTLS too.
1876 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1878 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1880 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1881 Applied second patch from Janne, implementing suggestion to default
1882 multiple-strings-in-record handling to match SPF spec.
1884 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1886 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1887 read-only, out of scope).
1888 Patch from Wolfgang Breyha, report from Stuart Northfield.
1890 PP/29 Fix three issues highlighted by clang analyser static analysis.
1891 Only crash-plausible issue would require the Cambridge-specific
1892 iplookup router and a misconfiguration.
1893 Report from Marcin Mirosław.
1895 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1897 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1898 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1899 As part of this, removing so much warning spew let me fix some minor
1900 real issues in debug logging.
1902 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1903 assignment on my part. Fixed.
1905 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1906 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1907 Janne Snabb (who went above and beyond: thank you).
1909 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1910 string otherwise requires a connection and a bunch more work and it's
1911 relatively easy to get wrong. Should also expose TLS library linkage
1914 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1915 64-bit ${eval} (JH/03).
1917 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1918 GNU libc to support some of the 64-bit stuff, should not lead to
1919 conflicts. Defined before os.h is pulled in, so if a given platform
1920 needs to override this, it can.
1922 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1923 protection layer was required, which is not implemented.
1924 Bugzilla 1254, patch from Wolfgang Breyha.
1926 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1927 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1928 tls_dhparam take prime identifiers. Also unbreak combination of
1929 OpenSSL+DH_params+TLSSNI.
1931 PP/39 Disable SSLv2 by default in OpenSSL support.
1937 PP/01 Solaris build fix for Oracle's LDAP libraries.
1938 Bugzilla 1109, patch from Stephen Usher.
1940 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1942 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1945 TF/02 Fix a couple more cases where we did not log the error message
1946 when unlink() failed. See also change 4.74-TF/03.
1948 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1949 lock up or crash if it happened to be inside a call to libc when it
1950 got a SIGUSR1 from exiwhat.
1952 The SIGUSR1 handler appends the current process status to the process
1953 log which is later printed by exiwhat. It used to use the general
1954 purpose logging code to do this, but several functions it calls are
1955 not safe for signals.
1957 The new output code in the SIGUSR1 handler is specific to the process
1958 log, and simple enough that it's easy to inspect for signal safety.
1959 Removing some special cases also simplifies the general logging code.
1960 Removing the spurious timestamps from the process log simplifies
1963 TF/04 Improved ratelimit ACL condition.
1965 The /noupdate option has been deprecated in favour of /readonly which
1966 has clearer semantics. The /leaky, /strict, and /readonly update modes
1967 are mutually exclusive. The update mode is no longer included in the
1968 database key; it just determines when the database is updated. (This
1969 means that when you upgrade Exim will forget old rate measurements.)
1971 Exim now checks that the per_* options are used with an update mode that
1972 makes sense for the current ACL. For example, when Exim is processing a
1973 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1974 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1975 must specify per_mail/readonly. If you omit the update mode it defaults to
1976 /leaky where that makes sense (as before) or /readonly where required.
1978 The /noupdate option is now undocumented but still supported for
1979 backwards compatibility. It is equivalent to /readonly except that in
1980 ACLs where /readonly is required you may specify /leaky/noupdate or
1981 /strict/noupdate which are treated the same as /readonly.
1983 A useful new feature is the /count= option. This is a generalization
1984 of the per_byte option, so that you can measure the throughput of other
1985 aggregate values. For example, the per_byte option is now equivalent
1986 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1988 The per_rcpt option has been generalized using the /count= mechanism
1989 (though it's more complicated than the per_byte equivalence). When it is
1990 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1991 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1992 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1993 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1994 Note that using per_rcpt with a non-readonly update mode in more than
1995 one ACL will cause the recipients to be double-counted. (The per_mail
1996 and per_byte options don't have this problem.)
1998 The handling of very low rates has changed slightly. If the computed rate
1999 is less than the event's count (usually one) then this event is the first
2000 after a long gap. In this case the rate is set to the same as this event's
2001 count, so that the first message of a spam run is counted properly.
2003 The major new feature is a mechanism for counting the rate of unique
2004 events. The new per_addr option counts the number of different
2005 recipients that someone has sent messages to in the last time period. It
2006 behaves like per_rcpt if all the recipient addresses are different, but
2007 duplicate recipient addresses do not increase the measured rate. Like
2008 the /count= option this is a general mechanism, so the per_addr option
2009 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2010 example, measure the rate that a client uses different sender addresses
2011 with the options per_mail/unique=$sender_address. There are further
2012 details in the main documentation.
2014 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2016 TF/06 Removed a few PCRE remnants.
2018 TF/07 Automatically extract Exim's version number from tags in the git
2019 repository when doing development or release builds.
2021 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2022 Bugzilla 879. Patch from Paul Fisher.
2024 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2025 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2028 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2030 PP/05 Variable $av_failed, true if the AV scanner deferred.
2031 Bugzilla 1078. Patch from John Horne.
2033 PP/06 Stop make process more reliably on build failure.
2034 Bugzilla 1087. Patch from Heiko Schlittermann.
2036 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2037 Bugzilla 1089. Patch from Heiko Schlittermann.
2039 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2040 Bugzilla 1131. Patch from Holger Weiß.
2042 PP/09 Handle IPv6 addresses with SPF.
2043 Bugzilla 860. Patch from Wolfgang Breyha.
2045 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2047 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2050 PP/11 match_* no longer expand right-hand-side by default.
2051 New compile-time build option, EXPAND_LISTMATCH_RHS.
2052 New expansion conditions, "inlist", "inlisti".
2054 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2056 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2058 PP/14 fix log_write() format string regression from TF/03.
2059 Bugzilla 1152. Patch from Dmitry Isaikin.
2065 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2067 PP/02 Harmonised TLS library version reporting; only show if debugging.
2068 Layout now matches that introduced for other libraries in 4.74 PP/03.
2070 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2072 PP/04 New "dns_use_edns0" global option.
2074 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2077 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2078 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2080 TK/01 Updated PolarSSL code to 0.14.2.
2081 Bugzilla 1097. Patch from Andreas Metzler.
2083 PP/07 Catch divide-by-zero in ${eval:...}.
2084 Fixes bugzilla 1102.
2086 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2089 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2090 format-string attack -- SECURITY: remote arbitrary code execution.
2092 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2093 time unintentionally subject to list matching rules, letting the header
2094 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2095 arbitrary string expansion). This allowed for information disclosure.
2097 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2098 INT_MIN/-1 -- value coerced to INT_MAX.
2104 NM/01 Workaround for PCRE version dependency in version reporting
2107 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2108 This fixes portability to compilers other than gcc, notably
2109 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2111 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2112 makefiles for portability to HP-UX and POSIX correctness.
2114 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2115 Also via indented variable definition in the Makefile.
2116 (Debugging by Oliver Heesakkers).
2118 PP/02 Restore caching of spamd results with expanded spamd_address.
2119 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2121 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2122 Improves build reliability. Fix from: Frank Elsner
2124 NM/02 Fix wide character breakage in the rfc2047 coding
2125 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2127 NM/03 Allow underscore in dnslist lookups
2128 Fixes bug 1026. Patch from Graeme Fowler
2130 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2131 Code patches from Adam Ciarcinski of NetBSD.
2133 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2136 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2137 is logged, to avoid truncation. Patch from John Horne.
2139 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2140 Patch from Jakob Hirsch.
2142 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2143 SQL string expansion failure details.
2144 Patch from Andrey Oktyabrski.
2146 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2147 Patch from Simon Arlott.
2149 PP/09 New lookups functionality failed to compile on old gcc which rejects
2150 extern declarations in function scope.
2151 Patch from Oliver Fleischmann
2153 PP/10 Use sig_atomic_t for flags set from signal handlers.
2154 Check getgroups() return and improve debugging.
2155 Fixed developed for diagnosis in bug 927 (which turned out to be
2158 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2159 Patch from Mark Zealey.
2161 PP/12 Bugzilla 1056: Improved spamd server selection.
2162 Patch from Mark Zealey.
2164 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2165 Based on patch from Heiko Schlittermann.
2167 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2168 Patch from Uwe Doering, sign-off by Michael Haardt.
2170 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2171 variable declaration deep within a block. Bug and patch from
2174 PP/15 lookups-Makefile IRIX compatibility coercion.
2176 PP/16 Make DISABLE_DKIM build knob functional.
2178 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2179 Patch by Simon Arlott
2181 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2182 variable argument macros. Our copy now differs from upstream.
2188 TF/01 Failure to get a lock on a hints database can have serious
2189 consequences so log it to the panic log.
2191 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2192 controlled using the smtp_confirmation log selector.
2194 TF/03 Include the error message when we fail to unlink a spool file.
2196 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2197 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2198 for maintaining out-of-tree patches for some time.
2200 PP/01 Bugzilla 139: Documentation and portability issues.
2201 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2202 Handle per-OS dynamic-module compilation flags.
2204 PP/02 Let /dev/null have normal permissions.
2205 The 4.73 fixes were a little too stringent and complained about the
2206 permissions on /dev/null. Exempt it from some checks.
2207 Reported by Andreas M. Kirchwitz.
2209 PP/03 Report version information for many libraries, including
2210 Exim version information for dynamically loaded libraries. Created
2211 version.h, now support a version extension string for distributors
2212 who patch heavily. Dynamic module ABI change.
2214 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2215 privilege escalation vulnerability whereby the Exim run-time user
2216 can cause root to append content of the attacker's choosing to
2219 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2222 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2223 If dropping privileges for untrusted macros, we disabled normal logging
2224 on the basis that it would fail; for the Exim run-time user, this is not
2225 the case, and it resulted in successful deliveries going unlogged.
2226 Fixed. Reported by Andreas Metzler.
2232 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2233 only prepend for the Resent-* case. Fixes regression introduced in
2234 Exim 4.70 by NM/22 for Bugzilla 607.
2236 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2237 increasing numbers of administrators be bitten by this.
2239 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2241 PP/03 Bugzilla 994: added openssl_options main configuration option.
2243 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2245 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2247 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2249 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2250 without a peer certificate, leading to a segfault because of an
2251 assumption that peers always have certificates. Be a little more
2252 paranoid. Problem reported by Martin Tscholak.
2254 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2255 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2256 NB: ClamAV planning to remove STREAM in "middle of 2010".
2257 CL also introduces -bmalware, various -d+acl logging additions and
2258 more caution in buffer sizes.
2260 PP/09 Implemented reverse_ip expansion operator.
2262 PP/10 Bugzilla 937: provide a "debug" ACL control.
2264 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2266 PP/12 Bugzilla 973: Implement --version.
2268 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2270 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2272 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2274 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2275 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2276 ignore trailing whitespace.
2278 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2280 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2283 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2284 Notification from Dr Andrew Aitchison.
2286 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2287 ExtendedDetectionInfo response format.
2288 Notification from John Horne.
2290 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2293 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2294 XSL and documented dependency on system catalogs, with examples of how
2297 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2300 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2301 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2302 configuration file which is writeable by the Exim user or group.
2304 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2305 of configuration files to cover files specified with the -C option if
2306 they are going to be used with root privileges, not just the default
2309 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2310 option (effectively making it always true).
2312 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2313 files to be used while preserving root privileges.
2315 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2316 that rogue child processes cannot use them.
2318 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2319 run-time user, instead of root.
2321 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2322 Exim run-time user without dropping privileges.
2324 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2325 result string, instead of calling string_vformat() twice with the same
2328 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2329 for other users. Others should always drop root privileges if they use
2330 -C on the command line, even for a whitelisted configure file.
2332 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2334 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2340 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2341 $data_path, and $header_path variables; fixed documentation bugs and
2344 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2345 exipick to access non-standard spools, including the "frozen" queue
2348 NM/01 Bugzilla 965: Support mysql stored procedures.
2349 Patch from Alain Williams
2351 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2353 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2354 Patch from Andreas Metzler
2356 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2357 Patch from Kirill Miazine
2359 NM/05 Bugzilla 671: Added umask to procmail example.
2361 JJ/03 installed exipick 20100323.0, fixing doc bug
2363 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2364 directory. Notification and patch from Dan Rosenberg.
2366 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2368 TK/02 Improve log output when DKIM signing operation fails.
2370 MH/01 Treat the transport option dkim_domain as a colon separated
2371 list, not as a single string, and sign the message with each element,
2372 omitting multiple occurences of the same signer.
2374 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2375 Bugzilla 985, 986. Patch by Simon Arlott
2377 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2378 Patch by Simon Arlott
2380 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2381 MBX locking. Notification from Dan Rosenberg.
2387 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2389 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2391 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2393 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2395 NM/04 Bugzilla 915: Flag broken perl installation during build.
2401 TK/01 Added patch by Johannes Berg that expands the main option
2402 "spamd_address" if it starts with a dollar sign.
2404 TK/02 Write list of recipients to X-Envelope-Sender header when building
2405 the mbox-format spool file for content scanning (suggested by Jakob
2408 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2409 (http://www.dcc-servers.net/) support via dccifd. Activated by
2410 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2412 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2413 by Mark Daniel Reidel <mr@df.eu>.
2415 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2416 When building exim an external PCRE library is now needed -
2417 PCRE is a system library on the majority of modern systems.
2418 See entry on PCRE_LIBS in EDITME file.
2420 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2421 conversation. Added nologin parameter to request.
2422 Patch contributed by Kirill Miazine.
2424 TF/01 Do not log submission mode rewrites if they do not change the address.
2426 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2428 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2429 log files in place. Contributed by Roberto Lima.
2431 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2433 TF/03 Bugzilla 615: When checking the local_parts router precondition
2434 after a local_part_suffix or local_part_prefix option, Exim now
2435 does not use the address's named list lookup cache, since this
2436 contains cached lookups for the whole local part.
2438 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2439 Robert Millan. Documentation is in experimental-spec.txt.
2441 TF/04 Bugzilla 668: Fix parallel build (make -j).
2443 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2445 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2446 Patch provided by Jan Srzednicki.
2448 TF/05 Leading white space used to be stripped from $spam_report which
2449 wrecked the formatting. Now it is preserved.
2451 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2452 that they are available at delivery time.
2454 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2456 TF/08 TLS error reporting now respects the incoming_interface and
2457 incoming_port log selectors.
2459 TF/09 Produce a more useful error message if an SMTP transport's hosts
2460 setting expands to an empty string.
2462 NM/06 Bugzilla 744: EXPN did not work under TLS.
2463 Patch provided by Phil Pennock.
2465 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2466 Patch provided by Richard Godbee.
2468 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2469 acl_smtp_notquit, added index entry.
2471 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2472 Patch provided by Eugene Bujak.
2474 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2475 accept(). Patch provided by Maxim Dounin.
2477 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2478 Patch provided by Phil Pennock.
2480 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2482 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2483 Patch provided by Brad "anomie" Jorsch.
2485 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2486 Patch provided by Dean Brooks.
2488 TK/05 Add native DKIM support (does not depend on external libraries).
2490 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2491 Patch provided by Graeme Fowler.
2493 NM/16 Bugzilla 851: Documentation example syntax fix.
2495 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2497 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2500 NM/19 Bugzilla 745: TLS version reporting.
2501 Patch provided by Phil Pennock.
2503 NM/20 Bugzilla 167: bool: condition support.
2504 Patch provided by Phil Pennock.
2506 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2507 clients. Patch provided by Phil Pennock.
2509 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2510 Patch provided by Brad "anomie" Jorsch.
2512 NM/23 Bugzilla 687: Fix misparses in eximstats.
2513 Patch provided by Heiko Schlittermann.
2515 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2516 Patch provided by Heiko Schlittermann.
2518 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2519 plus update to original patch.
2521 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2523 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2524 Patch provided by David Brownlee.
2526 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2528 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2530 NM/30 Bugzilla 888: TLS documentation bugfixes.
2532 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2534 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2535 Unlike the original bugzilla I have changed all shell scripts in src tree.
2537 NM/33 Bugzilla 898: Transport filter timeout fix.
2538 Patch by Todd Rinaldo.
2540 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2541 Patch by Serge Demonchaux.
2543 NM/35 Bugzilla 39: Base64 decode bug fixes.
2544 Patch by Jakob Hirsch.
2546 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2548 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2550 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2552 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2558 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2559 ALT-N's libdkim that I have put here:
2560 http://duncanthrax.net/exim-experimental/
2562 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2563 were called 'true' and it seems that C99 defines that as a reserved
2564 keyword to be used with 'bool' variable types. That means you could
2565 not include C99-style headers which use bools without triggering
2566 build errors in sieve.c.
2568 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2569 as mailq or other aliases. Changed the --help handling significantly
2570 to do whats expected. exim_usage() emits usage/help information.
2572 SC/01 Added the -bylocaldomain option to eximstats.
2574 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2576 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2578 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2584 PH/01 Another patch from the Sieve maintainer.
2586 PH/02 When an IPv6 address is converted to a string for single-key lookup
2587 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2588 dots are used instead of colons so that keys in lsearch files need not
2589 contain colons. This was done some time before quoting was made available
2590 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2591 (notated using the quote facility) so as to distinguish them from IPv4
2592 keys. This meant that lookups for IP addresses in host lists did not work
2593 for iplsearch lookups.
2595 This has been fixed by arranging for IPv6 addresses to be expressed with
2596 colons if the lookup type is iplsearch. This is not incompatible, because
2597 previously such lookups could never work.
2599 The situation is now rather anomalous, since one *can* have colons in
2600 ordinary lsearch keys. However, making the change in all cases is
2601 incompatible and would probably break a number of configurations.
2603 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2606 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2607 conversion specification without a maximum field width, thereby enabling
2608 a rogue spamd server to cause a buffer overflow. While nobody in their
2609 right mind would setup Exim to query an untrusted spamd server, an
2610 attacker that gains access to a server running spamd could potentially
2611 exploit this vulnerability to run arbitrary code as the Exim user.
2613 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2614 $primary_hostname instead of what libspf2 thinks the hosts name is.
2616 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2617 a directory entry by the name of the lookup key. Previously, if a
2618 symlink pointed to a non-existing file or a file in a directory that
2619 Exim lacked permissions to read, a lookup for a key matching that
2620 symlink would fail. Now it is enough that a matching directory entry
2621 exists, symlink or not. (Bugzilla 503.)
2623 PH/03 The body_linecount and body_zerocount variables are now exported in the
2626 PH/04 Added the $dnslist_matched variable.
2628 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2629 This means they are set thereafter only if the connection becomes
2632 PH/06 Added the client_condition to authenticators so that some can be skipped
2633 by clients under certain conditions.
2635 PH/07 The error message for a badly-placed control=no_multiline_responses left
2636 "_responses" off the end of the name.
2638 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2640 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2641 (without spaces) instead of just copying the configuration text.
2643 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2645 PH/11 Added $max_received_linelength.
2647 PH/12 Added +ignore_defer and +include_defer to host lists.
2649 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2650 way in which PCRE > 7.0 is built.
2652 PH/14 Implemented queue_only_load_latch.
2654 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2655 MAIL command. The effect was to mangle the value on 64-bit systems.
2657 PH/16 Another patch from the Sieve maintainer.
2659 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2661 PH/18 If a system quota error occurred while trying to create the file for
2662 a maildir delivery, the message "Mailbox is full" was not appended to the
2663 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2664 only to a quota excession during the actual writing of the file.
2666 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2667 characters?) which causes problems in log lines. The DN values are now
2668 passed through string_printing() before being added to log lines.
2670 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2671 and InterBase are left for another time.)
2673 PH/21 Added message_body_newlines option.
2675 PH/22 Guard against possible overflow in moan_check_errorcopy().
2677 PH/23 POSIX allows open() to be a macro; guard against that.
2679 PH/24 If the recipient of an error message contained an @ in the local part
2680 (suitably quoted, of course), incorrect values were put in $domain and
2681 $local_part during the evaluation of errors_copy.
2687 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2688 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2691 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2692 issue a MAIL command.
2694 PH/02 In an ACL statement such as
2696 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2698 if a client was not listed at all, or was listed with a value other than
2699 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2700 the condition was not true (as it should be), so access was not denied.
2701 The bug was that the ! inversion was incorrectly passed on to the second
2702 item. This has been fixed.
2704 PH/03 Added additional dnslists conditions == and =& which are different from
2705 = and & when the dns lookup returns more than one IP address.
2707 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2708 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2710 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2711 FSYNC, which compiles an option called disable_fsync that allows for
2712 bypassing fsync(). The documentation is heavily laced with warnings.
2714 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2716 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2717 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2718 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2719 including adding "make clean"; (3) Added -fPIC when compiling the test
2720 dynamically loaded module, to get rid of a warning.
2722 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2723 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2724 = 0s. The bug is otherwise harmless.
2726 PH/07 There was a bug in the dovecot authenticator such that the value of
2727 $auth1 could be overwritten, and so not correctly preserved, after a
2728 successful authentication. This usually meant that the value preserved by
2729 the server_setid option was incorrect.
2731 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2733 PH/09 Installed PCRE release 7.0.
2735 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2736 run for batched SMTP input. It is now run at the start of every message
2737 in the batch. While fixing this I discovered that the process information
2738 (output by running exiwhat) was not always getting set for -bs and -bS
2739 input. This is fixed, and it now also says "batched" for BSMTP.
2741 PH/11 Added control=no_pipelining.
2743 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2744 patch, slightly modified), and move the expansion of helo_data till after
2745 the connection is made in the smtp transport (so it can use these
2748 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2750 PH/14 Added log_selector = +pid.
2752 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2754 PH/16 Add ${if forany and ${if forall.
2756 PH/17 Added dsn_from option to vary the From: line in DSNs.
2758 PH/18 Flush SMTP output before performing a callout, unless control =
2759 no_callout_flush is set.
2761 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2762 was true (the default) a successful delivery failed to delete the retry
2763 item, thus causing premature timeout of the address. The bug is now
2766 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2768 PH/21 Long custom messages for fakedefer and fakereject are now split up
2769 into multiline responses in the same way that messages for "deny" and
2770 other ACL rejections are.
2772 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2773 with slight modification.
2775 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2776 draft, changing the syntax and factoring some duplicate code".
2778 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2779 for deliveries of the second and subsequent messages over the same SMTP
2782 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2783 ${reduce, with only minor "tidies".
2785 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2787 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2788 expansion side effects.
2790 PH/27 When a message times out after an over-quota error from an Exim-imposed
2791 quota, the bounce message says "mailbox is full". This message was not
2792 being given when it was a system quota that was exceeded. It now should
2795 MH/03 Made $recipients available in local_scan(). local_scan() already has
2796 better access to the recipient list through recipients_list[], but
2797 $recipients can be useful in postmaster-provided expansion strings.
2799 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2800 in the case of a MAIL command with additional options following the
2801 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2802 were accidentally chopped off.
2804 PH/29 SMTP synchronization checks are implemented when a command is read -
2805 there is a check that no more input is waiting when there shouldn't be
2806 any. However, for some commands, a delay in an ACL can mean that it is
2807 some time before the response is written. In this time, more input might
2808 arrive, invalidly. So now there are extra checks after an ACL has run for
2809 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2810 pipelining has not been advertised.
2812 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2814 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2815 returned in the SMTP response when smtp_return_error_details was false.
2816 This has been fixed.
2818 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2819 instead of the C I/O that was originally supplied, because problems were
2820 reported on Solaris.
2822 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2823 Exim which did not show up earlier: it was assuming that a call to
2824 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2825 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2826 was a macro that became an assignment, so it seemed to work. This has
2827 changed to a proper function call with a void return, hence the compile
2828 error. Exim's code has been fixed.
2830 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2833 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2835 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2837 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2838 because it turns out to be too big or there is a timeout) there is no
2839 "Completed" line in the log. When some messages of this type were
2840 selected by exigrep, they were listed as "not completed". Others were
2841 picked up by some special patterns. I have improved the selection
2842 criteria to be more general.
2844 PH/38 The host_find_failed option in the manualroute router can now be set
2845 to "ignore", to completely ignore a host whose IP address cannot be
2846 found. If all hosts are ignored, the behaviour is controlled by the new
2847 host_all_ignored option.
2849 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2850 homing or because of multiple MX records with /mx) generated more than
2851 one IP address, and the following item turned out to be the local host,
2852 all the secondary addresses of the first item were incorrectly removed
2853 from the list, along with the local host and any following hosts (which
2854 is what is supposed to happen).
2856 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2857 whoever called Exim into the -H file. In the case of the daemon it was
2858 behaving confusingly. When first started, it used values for whoever
2859 started the daemon, but after a SIGHUP it used the Exim user (because it
2860 calls itself on a restart). I have changed the code so that it now always
2863 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2864 message are rejected with the same error (e.g. no authentication or bad
2865 sender address), and a DATA command is nevertheless sent (as can happen
2866 with PIPELINING or a stupid MUA), the error message that was given to the
2867 RCPT commands is included in the rejection of the DATA command. This is
2868 intended to be helpful for MUAs that show only the final error to their
2871 PH/42 Another patch from the Sieve maintainer.
2873 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2874 Eximstats - Fixed some broken HTML links and added missing column headers
2876 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2877 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2879 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2881 PH/43 Yet another patch from the Sieve maintainer.
2883 PH/44 I found a way to check for a TCP/IP connection going away before sending
2884 the response to the final '.' that terminates a message, but only in the
2885 case where the client has not sent further data following the '.'
2886 (unfortunately, this is allowed). However, in many cases there won't be
2887 any further data because there won't be any more messages to send. A call
2888 to select() can be used: if it shows that the input is "ready", there is
2889 either input waiting, or the socket has been closed. An attempt to read
2890 the next input character can distinguish the two cases. Previously, Exim
2891 would have sent an OK response which the client would never have see.
2892 This could lead to message repetition. This fix should cure that, at
2893 least in a lot of common cases.
2895 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2896 advertised in response to EHLO.
2902 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2903 fixed by 4.65/MH/01 (is this a record?) are fixed:
2905 (i) An empty string was always treated as zero by the numeric comparison
2906 operators. This behaviour has been restored.
2908 (ii) It is documented that the numeric comparison operators always treat
2909 their arguments as decimal numbers. This was broken in that numbers
2910 starting with 0 were being interpreted as octal.
2912 While fixing these problems I realized that there was another issue that
2913 hadn't been noticed. Values of message_size_limit (both the global option
2914 and the transport option) were treated as octal if they started with 0.
2915 The documentation was vague. These values are now always treated as
2916 decimal, and I will make that clear in the documentation.
2922 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2923 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2926 MH/01 Don't check that the operands of numeric comparison operators are
2927 integers when their expansion is in "skipping" mode (fixes bug
2928 introduced by 4.64-PH/07).
2930 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2931 child addresses, Exim now panics and dies. Previously, because the count
2932 is held in a short int, deliveries were likely to be lost. As such a
2933 large number of recipients for a single message is ridiculous
2934 (performance will be very, very poor), I have chosen to impose a limit
2935 rather than extend the field.
2941 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2942 leftover -K file (the existence of which was triggered by #402).
2943 While we were at it, introduced process PID as part of the -K
2944 filename. This should rule out race conditions when creating
2947 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2948 processing considerably. Previous code took too long for large mails,
2949 triggering a timeout which in turn triggers #401.
2951 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2952 in the DK code in transports.c. sendfile() is not really portable,
2953 hence the _LINUX specificness.
2955 TF/01 In the add_headers option to the mail command in an Exim filter,
2956 there was a bug that Exim would claim a syntax error in any
2957 header after the first one which had an odd number of characters
2960 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2961 callout verification, Exim cached a "reject" for the entire domain. This
2962 is correct for most verifications, but it is not correct for a recipient
2963 verification with use_sender or use_postmaster set, because in that case
2964 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2965 case of MAIL FROM:<> rejection from other early rejections (e.g.
2966 rejection of HELO). When verifying a recipient using a non-null MAIL
2967 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2968 Whatever the result of the callout, the value of the domain cache is
2969 left unchanged (for any other kind of callout, getting as far as trying
2970 RCPT means that the domain itself is ok).
2972 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2975 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2976 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2977 session, the daemon ignores ECONNECT errors and logs others; it now
2978 ignores EPIPE as well.
2980 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2981 (quoted-printable decoding).
2983 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2984 later the small subsequent patch to fix an introduced bug.
2986 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2988 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2990 PH/08 An error is now given if message_size_limit is specified negative.
2992 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2993 to be given (somewhat) arbitrary names.
2995 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2998 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2999 miscellaneous code fixes
3001 PH/10 Added the log_reject_target ACL modifier to specify where to log
3004 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3005 hostname. This is wrong, because it relates to the incoming message (and
3006 probably the interface on which it is arriving) and not to the outgoing
3007 callout (which could be using a different interface). This has been
3008 changed to use the value of the helo_data option from the smtp transport
3009 instead - this is what is used when a message is actually being sent. If
3010 there is no remote transport (possible with a router that sets up host
3011 addresses), $smtp_active_hostname is used.
3013 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3014 tweaks were necessary in order to get it to work (see also 21 below):
3015 (a) The code assumed that strncpy() returns a negative number on buffer
3016 overflow, which isn't the case. Replaced with Exim's string_format()
3018 (b) There were several signed/unsigned issues. I just did the minimum
3019 hacking in of casts. There is scope for a larger refactoring.
3020 (c) The code used strcasecmp() which is not a standard C function.
3021 Replaced with Exim's strcmpic() function.
3022 (d) The code set only $1; it now sets $auth1 as well.
3023 (e) A simple test gave the error "authentication client didn't specify
3024 service in request". It would seem that Dovecot has changed its
3025 interface. Fortunately there's a specification; I followed it and
3026 changed what the client sends and it appears to be working now.
3028 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3031 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3032 address A is aliased to B and C, where B exists and C does not. Without
3033 -v the output is "A verified" because verification stops after a
3034 successful redirection if more than one address is generated. However,
3035 with -v the child addresses are also verified. Exim was outputting "A
3036 failed to verify" and then showing the successful verification for C,
3037 with its parentage. It now outputs "B failed to verify", showing B's
3038 parentage before showing the successful verification of C.
3040 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3041 look up a TXT record in a specific list after matching in a combined
3044 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3045 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3046 they consult the DNS. I had assumed they would set it the way they
3047 wanted; and indeed my experiments on Linux seem to show that in some
3048 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3049 To be on the safe side, however, I have now made the interface to
3050 host_find_byname() similar to host_find_bydns(), with an argument
3051 containing the DNS resolver options. The host_find_byname() function now
3052 sets these options at its start, just as host_find_bydns() does. The smtp
3053 transport options dns_qualify_single and dns_search_parents are passed to
3054 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3055 of host_find_byname() use the default settings of RES_DEFNAMES
3056 (qualify_single) but not RES_DNSRCH (search_parents).
3058 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3059 spool_read_header() do less string testing, by means of a preliminary
3060 switch on the second character of optional "-foo" lines. (This is
3061 overdue, caused by the large number of possibilities that now exist.
3062 Originally there were few.) While I was there, I also converted the
3063 str(n)cmp tests so they don't re-test the leading "-" and the first
3064 character, in the hope this might squeeze out yet more improvement.
3066 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3067 flag allowing group syntax was set by the header_syntax check but not
3068 turned off, possible causing trouble later; (2) The flag was not being
3069 set at all for the header_verify test, causing "group"-style headers to
3070 be rejected. I have now set it in this case, and also caused header_
3071 verify to ignore an empty address taken from a group. While doing this, I
3072 came across some other cases where the code for allowing group syntax
3073 while scanning a header line wasn't quite right (mostly, not resetting
3074 the flag correctly in the right place). These bugs could have caused
3075 trouble for malformed header lines. I hope it is now all correct.
3077 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3078 with the "reply" argument non-NULL. The code, however (which originally
3079 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3080 but it didn't always do it. This confused somebody who was copying the
3081 code for some other use. I have removed all the tests.
3083 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3084 feature that was used to support insecure browsers during the U.S. crypto
3085 embargo. It requires special client support, and Exim is probably the
3086 only MTA that supported it -- and would never use it because real RSA is
3087 always available. This code has been removed, because it had the bad
3088 effect of slowing Exim down by computing (never used) parameters for the
3089 RSA_EXPORT functionality.
3091 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3092 authenticator to fail if there's a tab character in the incoming data
3093 (there should never be unless someone is messing about, as it's supposed
3094 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3095 if the connection is using TLS or if the remote IP is the same as the
3096 local IP, and the "valid-client-cert option" if a client certificate has
3099 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3100 authenticators. This can be used for authorization after authentication
3101 succeeds. (In the case of plaintext, it servers for both authentication
3104 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3105 if any retry times were supplied.
3107 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3108 connection, where there is no client IP address to check. In this
3109 situation, the verify now always succeeds.
3111 PH/25 Applied John Jetmore's -Mset patch.
3113 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3115 PH/27 In a string expansion for a processed (not raw) header when multiple
3116 headers of the same name were present, leading whitespace was being
3117 removed from all of them, but trailing whitespace was being removed only
3118 from the last one. Now trailing whitespace is removed from each header
3119 before concatenation. Completely empty headers in a concatenation (as
3120 before) are ignored.
3122 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3123 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3125 PH/29 [Removed. This was a change that I later backed out, and forgot to
3126 correct the ChangeLog entry (that I had efficiently created) before
3127 committing the later change.]
3129 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3130 address errors (4xx response to RCPT) over the same connection as other
3131 messages routed to the same hosts. Such deliveries are always "forced",
3132 so retry times are not inspected. This resulted in far too many retries
3133 for the affected addresses. The effect occurred only when there were more
3134 hosts than the hosts_max_try setting in the smtp transport when it had
3135 the 4xx errors. Those hosts that it had tried were not added to the list
3136 of hosts for which the message was waiting, so if all were tried, there
3137 was no problem. Two fixes have been applied:
3139 (i) If there are any address or message errors in an SMTP delivery, none
3140 of the hosts (tried or untried) are now added to the list of hosts
3141 for which the message is waiting, so the message should not be a
3142 candidate for sending over the same connection that was used for a
3143 successful delivery of some other message. This seems entirely
3144 reasonable: after all the message is NOT "waiting for some host".
3145 This is so "obvious" that I'm not sure why it wasn't done
3146 previously. Hope I haven't missed anything, but it can't do any
3147 harm, as the worst effect is to miss an optimization.
3149 (ii) If, despite (i), such a delivery is accidentally attempted, the
3150 routing retry time is respected, so at least it doesn't keep
3151 hammering the server.
3153 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3154 in ${readsocket because some servers need this prod.
3156 PH/32 Added some extra debug output when updating a wait-xxx database.
3158 PH/33 The hint "could be header name not terminated by colon", which has been
3159 given for certain expansion errors for a long time, was not being given
3160 for the ${if def:h_colon_omitted{... case.
3162 PH/34 The spec says: "With one important exception, whenever a domain list is
3163 being scanned, $domain contains the subject domain." There was at least
3164 one case where this was not true.
3166 PH/35 The error "getsockname() failed: connection reset by peer" was being
3167 written to the panic log as well as the main log, but it isn't really
3168 panic-worthy as it just means the connection died rather early on. I have
3169 removed the panic log writing for the ECONNRESET error when getsockname()
3172 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3173 runs only) independently of the message's sender address. This meant
3174 that, if the 4xx error was in fact related to the sender, a different
3175 message to the same recipient with a different sender could confuse
3176 things. In particular, this can happen when sending to a greylisting
3177 server, but other circumstances could also provoke similar problems.
3178 I have changed the default so that the retry time for these errors is now
3179 based a combination of the sender and recipient addresses. This change
3180 can be overridden by setting address_retry_include_sender=false in the
3183 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3184 remote server are returned as part of bounce messages. This was not
3185 happening for LMTP over a pipe (the lmtp transport), but now it is the
3186 same for both kinds of LMTP.
3188 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3189 in header lines that were in fact CNAMEs. This is no longer the case.
3191 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3192 and queue runs started by the daemon processed all messages. This has
3193 been fixed so that -R and -S can now usefully be given with -q<time>.
3195 PH/40 Import PCRE release 6.7 (fixes some bugs).
3197 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3199 PH/42 Give an error if -q is specified more than once.
3201 PH/43 Renamed the variables $interface_address and $interface_port as
3202 $received_ip_address and $received_port, to make it clear that these
3203 values apply to message reception, and not to the outgoing interface when
3204 a message is delivered. (The old names remain recognized, of course.)
3206 PH/44 There was no timeout on the connect() call when using a Unix domain
3207 socket in the ${readsocket expansion. There now is.
3209 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3210 be meaningful with "accept".
3212 SC/01 Eximstats V1.43
3213 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3215 SC/02 Eximstats V1.44
3216 Use a glob alias rather than an array ref in the generated
3217 parser. This improves both readability and performance.
3219 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3220 Collect SpamAssassin and rejection statistics.
3221 Don't display local sender or destination tables unless
3222 there is data to show.
3223 Added average volumes into the top table text output.
3225 SC/04 Eximstats V1.46
3226 Collect data on the number of addresses (recipients)
3227 as well as the number of messages.
3229 SC/05 Eximstats V1.47
3230 Added 'Message too big' to the list of mail rejection
3231 reasons (thanks to Marco Gaiarin).
3233 SC/06 Eximstats V1.48
3234 Mainlog lines which have GMT offsets and are too short to
3235 have a flag are now skipped.
3237 SC/07 Eximstats V1.49 (Alain Williams)
3238 Added the -emptyok flag.
3240 SC/08 Eximstats V1.50
3241 Fixes for obtaining the IP address from reject messages.
3243 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3244 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3245 whitespace changes from 4.64-PH/27
3247 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3250 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3251 are found, allow negative numbers in numeric criteria)
3253 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3255 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3256 to match changes made in 4.64-PH/43
3258 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3260 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3261 "domain" is actually an IP address, and force a failure. This locks out
3262 those revolvers/nameservers that support "A-for-A" lookups, in
3263 contravention of the specifications.
3265 PH/48 When a host name was looked up from an IP address, and the subsequent
3266 forward lookup of the name timed out, the host name was left in
3267 $sender_host_name, contrary to the specification.
3269 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3270 restricted to single-key lookups, Exim was not diagnosing an error if
3271 * or *@ was used with a query-style lookup.
3273 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3275 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3276 long ago, but noone interested enough thought of it. Let's just say that
3277 the "1.1" means that there are some new functions that weren't there at
3278 some point in the past.
3280 PH/51 Error processing for expansion failure of helo_data from an smtp
3281 transport during callout processing was broken.
3283 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3284 tested/used via the -bh/-bhc/-bs options.
3286 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3287 bug, fixed in subsequent PCRE releases).
3289 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3290 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3292 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3298 SC/01 Use a glob alias rather than an array ref in eximstats generated
3299 parser. This improves both readability and performance.
3301 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3302 Don't display local sender or destination tables in eximstats unless
3303 there is data to show.
3304 Added average volumes into the eximstats top table text output.
3306 SC/03 Collect data on the number of addresses (recipients) as well
3307 as the number of messages in eximstats.
3309 TF/01 Correct an error in the documentation for the redirect router. Exim
3310 does (usually) call initgroups() when daemonizing.
3312 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3313 with consistent privilege compared to when running as a daemon.
3315 TF/03 Note in the spec that $authenticated_id is not set for local
3316 submissions from trusted users.
3318 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3319 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3321 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3322 by adding some example configuration directives to the default
3323 configuration file. A little bit of work is required to uncomment the
3324 directives and define how usernames and passwords are checked, but
3325 there is now a framework to start from.
3327 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3328 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3329 without this. I don't know how relevant this is to other LDAP libraries.
3331 PH/02 Add the verb name to the "unknown ACL verb" error.
3333 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3335 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3337 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3338 directory not expanded when it should be if an expanded home directory
3339 was set for the address (which is overridden by the transport).
3341 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3344 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3345 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3346 because it is too late at that time, and has no effect.
3348 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3349 security issue with \' (bugzilla #107). I could not use the
3350 PQescapeStringConn() function, because it needs a PGconn value as one of
3353 PH/08 When testing addresses using -bt, indicate those final addresses that
3354 are duplicates that would not cause an additional delivery. At least one
3355 person was confused, thinking that -bt output corresponded to deliveries.
3356 (Suppressing duplicates isn't a good idea as you lose the information
3357 about possibly different redirections that led to the duplicates.)
3359 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3360 systems where poll() doesn't work, in particular OS X.
3362 PH/10 Added more information to debugging output for retry time not reached.
3364 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3365 operations in malware.c.
3367 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3370 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3371 syslog_duplication set false, log lines that would normally be written
3372 both the the main log and to the reject log were not written to syslog at
3375 PH/14 In the default configuration, change the use of "message" in ACL warn
3376 statements to "add_header".
3378 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3379 not followed by a command (e.g. "seen endif").
3381 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3382 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3385 PH/17 Added extra conditions to the default value of delay_warning_condition
3389 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3390 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3391 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3394 The Auto-Submitted: and various List- headers are standardised, whereas I
3395 don't think Precedence: ever was.
3397 PH/18 Refactored debugging code in route_finduser() to show more information,
3398 in particular, the error code if getpwnam() issues one.
3400 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3401 This is apparently needed in addition to the PH/07 change above to avoid
3402 any possible encoding problems.
3404 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3405 but not after initializing Perl.
3407 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3408 output them only if debugging. By default they are written stderr,
3409 apparently, which is not desirable.
3411 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3414 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3417 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3419 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3420 authenticated or an ident call has been made. Suppress the default
3421 values for $authenticated_id and $authenticated_sender (but permit -oMai
3422 and -oMas) when testing with -bh.
3424 PH/24 Re-jigged the order of the tests in the default configuration so that the
3425 tests for valid domains and recipients precede the DNS black list and CSA
3426 tests, on the grounds that those ones are more expensive.
3428 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3429 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3430 This bug exists in every version of Exim that I still have, right back to
3433 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3434 However, an attempt to turn on case-sensitivity in a regex key by
3435 including (?-i) didn't work because the subject string was already
3436 lowercased, and the effects were non-intuitive. It turns out that a
3437 one-line patch can be used to allow (?-i) to work as expected.
3443 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3444 other effects) broke the use of negated acl sub-conditions.
3446 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3449 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3450 "Deny" causes Exim to reject the incoming connection with a 554 error.
3451 Unfortunately, if there is a major crisis, such as a disk failure,
3452 tcp-wrappers gives "deny", whereas what one would like would be some
3453 kind of temporary error. A kludge has been added to help with this.
3454 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3455 554 error is used if errno is still zero or contains ENOENT (which occurs
3456 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3459 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3461 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3462 errors. Otherwise a message that provokes a temporary error (when other
3463 messages do not) can cause a whole host to time out.
3465 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3466 addresses were routed directly to files or pipes from a redirect router.
3467 File deliveries just didn't batch; pipe deliveries might have suffered
3470 PH/06 A failure to get a lock for a hints database would erroneously always say
3471 "Failed to get write lock", even when it was really a read lock.
3473 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3474 of 0600. This has been changed to use the value of the lockfile_mode
3475 option (which defaults to 0600).
3477 PH/08 Applied small patch from the Sieve maintainer.
3479 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3480 folder from quota calculations, a direct delivery into this folder messed
3481 up the contents of the maildirsize file. This was because the regex was
3482 used only to exclude .Trash (or whatever) when the size of the mailbox
3483 was calculated. There was no check that a delivery was happening into an
3484 excluded directory. This bug has been fixed by ignoring all quota
3485 processing for deliveries into excluded directories.
3487 PH/10 Added the maildirfolder_create_regex option to appendfile.
3493 PH/01 The code for finding all the local interface addresses on a FreeBSD
3494 system running IPv6 was broken. This may well have applied to all BSD
3495 systems, as well as to others that have similar system calls. The broken
3496 code found IPv4 interfaces correctly, but gave incorrect values for the
3497 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3498 that it would not match correctly against @[] and not recognize the IPv6
3501 PH/02 The ipliteral router was not recognizing addresses of the form user@
3502 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3504 PH/03 Added disable_ipv6.
3506 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3507 decoded form, because it is most often used to construct To: headers
3508 lines in autoreplies, and the decoded form may well be syntactically
3509 invalid. However, $reply_address has leading white space removed, and all
3510 newlines turned into spaces so that the autoreply transport does not
3513 PH/05 If group was specified without a user on a router, and no group or user
3514 was specified on a transport, the group from the router was ignored.
3516 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3517 for visible compile-time settings that can be used to change these
3518 numbers, for those that want even more. Backwards compatibility with old
3519 spool files has been maintained. However, going back to a previous Exim
3520 release will lost any variables that are in spool files.
3522 PH/07 Two small changes when running in the test harness: increase delay when
3523 passing a TCP/IP connection to a new process, in case the original
3524 process has to generate a bounce, and remove special handling of
3525 127.0.0.2 (sic), which is no longer necessary.
3527 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3528 be the same on different OS.
3530 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3533 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3534 whether --show-vars was specified or not
3536 JJ/02 exipick: Added support for new ACL variable spool format introduced
3539 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3540 syntactically invalid From: or Reply-to: line, and a filter used this to
3541 generate an autoreply, and therefore failed to obtain an address for the
3542 autoreply, Exim could try to deliver to a non-existent relative file
3543 name, causing unrelated and misleading errors. What now happens is that
3544 it logs this as a hard delivery error, but does not attempt to create a
3547 PH/11 The exinext utility has a -C option for testing purposes, but although
3548 the given file was scanned by exinext itself; it wasn't being passed on
3549 when Exim was called.
3551 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3552 an end-of-file indication when reading a command response.
3554 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3555 compiled. In many other places in Exim, IPv6 addresses are always
3556 recognized, so I have changed this. It also means that IPv4 domain
3557 literals of the form [IPV4:n.n.n.n] are now always recognized.
3559 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3560 used if the router is not running as root, for example, when verifying at
3561 ACL time, or when using -bh. The debugging output from this situation was
3562 non-existent - all you got was a failure to exec. I have made two
3565 (a) Failures to set uid/gid, the current directory, or a process leader
3566 in a subprocess such as that created by queryprogram now generate
3567 suitable debugging output when -d is set.
3569 (b) The queryprogram router detects when it is not running as root,
3570 outputs suitable debugging information if -d is set, and then runs
3571 the subprocess without attempting to change uid/gid.
3573 PH/15 Minor change to Makefile for building test_host (undocumented testing
3576 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3577 additional section of a DNS packet that returns MX or SRV records.
3578 Instead, it always explicitly searches for A/AAAA records. This avoids
3579 major problems that occur when a DNS server includes only records of one
3580 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3581 fixed another bug: if SRV records were looked up and the corresponding
3582 address records were *not* found in the additional section, the port
3583 values from the SRV records were lost.
3585 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3586 using the correct key (the original address) when searching the retry
3587 rules in order to find which one to use for generating the retry hint.
3589 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3590 adding the default one. Similarly, if it contains a Reply-To: header, the
3591 errors_reply_to option, if set, is not used.
3593 PH/19 When calculating a retry time, Exim used to measure the "time since
3594 failure" by looking at the "first failed" field in the retry record. Now
3595 it does not use this if it is later than than the arrival time of the
3596 message. Instead it uses the arrival time. This makes for better
3597 behaviour in cases where some deliveries succeed, thus re-setting the
3598 "first failed" field. An example is a quota failure for a huge message
3599 when small messages continue to be delivered. Without this change, the
3600 "time since failure" will always be short, possible causing more frequent
3601 delivery attempts for the huge message than are intended.
3602 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3604 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3605 $1, $2, $3) because the numerical variables can be reset during some
3606 expansion items (e.g. "match"), thereby losing the authentication data.
3608 PH/21 Make -bV show the size of off_t variables so that the test suite can
3609 decide whether to run tests for quotas > 2G.
3611 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3612 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3613 filecount value is greater than 2G or if a quota value is greater than 2G
3614 on a system where the size of off_t is not greater than 4, a panic error
3617 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3618 never match. The debug and -bh output now contains an explicit error
3619 message indicating a malformed IPv4 address or mask.
3621 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3622 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3623 PH/23 above applies.
3625 PH/25 Do not write to syslog when running in the test harness. The only
3626 occasion when this arises is a failure to open the main or panic logs
3627 (for which there is an explicit test).
3629 PH/26 Added the /no_tell option to "control=freeze".
3631 PH/27 If a host name lookup failed very early in a connection, for example, if
3632 the IP address matched host_lookup and the reverse lookup yielded a name
3633 that did not have a forward lookup, an error message of the form "no IP
3634 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3635 could be logged. Now it outputs the IP address instead of "NULL".
3637 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3638 allows the sender and the authenticated sender to be set when
3639 submitting a message from within Exim. Since child_open_exim() is
3640 documented for local_scan(), the new function should be too.
3642 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3643 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3644 results in an empty string is now treated as unset.
3646 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3648 PH/31 Added sender_verify_fail logging option.
3650 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3651 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3652 tidied the source and removed it altogether.
3654 PH/33 When a queue run was abandoned because the load average was too high, a
3655 log line was always written; now it is written only if the queue_run log
3656 selector is set. In addition, the log line for abandonment now contains
3657 information about the queue run such as the pid. This is always present
3658 in "start" and "stop" lines but was omitted from the "abandon" line.
3660 PH/34 Omit spaces between a header name and the colon in the error message that
3661 is given when verify = headers_syntax fails (if there are lots of them,
3662 the message gets confusing).
3664 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3665 names, as there are now some PTR records that contain slashes. This check
3666 is only to protect against broken name servers that fall over on strange
3667 characters, so the fact that it applies to all lookups doesn't matter.
3669 PH/36 Now that the new test suite is complete, we can remove some of the
3670 special code in Exim that was needed for the old test suite. For example,
3671 sorting DNS records because real resolvers return them in an arbitrary
3672 order. The new test suite's fake resolver always returns records in the
3675 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3676 bounces) except when queue_only is set, to avoid logging races between
3677 the different processes.
3679 PH/38 Panic-die if .include specifies a non-absolute path.
3681 PH/39 A tweak to the "H" retry rule from its user.
3683 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3684 a label. They prevented compilation on older perls.
3686 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3687 a warning to be raised on newish perls.
3689 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3690 on queue. Changes to match documented behaviour of showing count of
3691 messages matching specified criteria.
3693 PH/40 Changed the default ident timeout from 30s to 5s.
3695 PH/41 Added support for the use of login_cap features, on those BSD systems
3696 that have them, for controlling the resources used by pipe deliveries.
3698 PH/42 The content-scanning code uses fopen() to create files in which to put
3699 message data. Previously it was not paying any attention to the mode of
3700 the files. Exim runs with umask(0) because the rest of the code creates
3701 files with open(), and sets the required mode explicitly. Thus, these
3702 files were ending up world-writeable. This was not a big issue, because,
3703 being within the spool directory, they were not world-accessible. I have
3704 created a function called modefopen, which takes an additional mode
3705 argument. It sets umask(777), creates the file, chmods it to the required
3706 mode, then resets the umask. All the relevant calls to fopen() in the
3707 content scanning code have been changed to use this function.
3709 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3710 to 24 hours. This avoids potential overflow problems when processing G
3711 and H retry rules. I suspect nobody ever tinkers with this value.
3713 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3715 PH/45 When the plaintext authenticator is running as a client, the server's
3716 challenges are checked to ensure they are valid base64 strings. By
3717 default, the authentication attempt is cancelled if an invalid string is
3718 received. Setting client_ignore_invalid_base64 true ignores these errors.
3719 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3720 they are received. Thus, the responses can be made to depend on the
3721 challenges. If an invalid string is ignored, an empty string is placed in
3724 PH/46 Messages that are created by the autoreply transport now contains a
3725 References: header, in accordance with RFCs 2822 and 3834.
3727 PH/47 Added authenticated_sender_force to the smtp transport.
3729 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3731 PH/49 Installed latest patch from the Sieve maintainer.
3733 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3734 was also set, the appendfile transport was unnecessarily scanning a
3735 directory of message files (e.g. for maildir delivery) to find the count
3736 of files (along with the size), even though it did not need this
3737 information. It now does the scan only if it needs to find either the
3738 size of the count of files.
3740 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3742 PH/52 Two bugs concerned with error handling when the smtp transport is
3745 (i) Exim was not creating retry information for temporary errors given
3746 for individual recipients after the DATA command when the smtp transport
3747 was used in LMTP mode. This meant that they could be retried too
3748 frequently, and not timed out correctly.
3750 (ii) Exim was setting the flag that allows error details to be returned
3751 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3752 recipients that were returned after the DATA command.
3754 PH/53 This is related to PH/52, but is more general: for any failing address,
3755 when detailed error information was permitted to be returned to the
3756 sender, but the error was temporary, then after the final timeout, only
3757 "retry timeout exceeded" was returned. Now it returns the full error as
3758 well as "retry timeout exceeded".
3760 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3761 do this, and (what is worse) MTAs that accept it.
3763 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3764 will now be deprecated.
3766 PH/56 New os.c-cygwin from the Cygwin maintainer.
3768 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3769 formats (previously only available in exim formats via -bpr, -bpru,
3770 and -bpra. Now also available in native and exiqgrep formats)
3772 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3773 with very large, slow to parse queues
3775 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3777 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3779 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3780 responses to authentication challenges, though it was showing the
3781 challenges; (ii) I've removed the CR characters from the debug output for
3784 PH/58 Allow for the insertion of a newline as well as a space when a string
3785 is turned into more than one encoded-word during RFC 2047 encoding. The
3786 Sieve code now uses this.
3788 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3789 data_4xx, lost_connection, tls_required.
3791 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3792 message was being sent as an SMTP response.
3794 PH/61 Add -l and -k options to exicyclog.
3796 PH/62 When verifying, if an address was redirected to one new address, so that
3797 verification continued, and the new address failed or deferred after
3798 having set something in $address_data, the value of $address_data was not
3799 passed back to the ACL. This was different to the case when no
3800 redirection occurred. The value is now passed back in both cases.
3802 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3803 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3804 use login_cap.h, so on its own it isn't the distinguishing feature. The
3805 new name refers directly to the setclassresources() function.
3807 PH/65 Added configuration files for NetBSD3.
3809 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3811 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3812 is preferred over IPv4.
3814 PH/68 The bounce_return_message and bounce_return_body options were not being
3815 honoured for bounces generated during the reception of non-SMTP messages.
3816 In particular, this applied to messages rejected by the ACL. This bug has
3817 been fixed. However, if bounce_return_message is true and bounce_return_
3818 body is false, the headers that are returned for a non-SMTP message
3819 include only those that have been read before the error was detected.
3820 (In the case of an ACL rejection, they have all been read.)
3822 PH/69 The HTML version of the specification is now built in a directory called
3823 spec_html instead of spec.html, because the latter looks like a path with
3824 a MIME-type, and this confuses some software.
3826 PH/70 Catch two compiler warnings in sieve.c.
3828 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3829 function verify_get_ident() calls ip_connect() to connect a socket, but
3830 if the "connect()" function timed out, ip_connect() used to close the
3831 socket. However, verify_get_ident() also closes the socket later, and in
3832 between Exim writes to the log, which may get opened at this point. When
3833 the socket was closed in ip_connect(), the log could get the same file
3834 descriptor number as the socket. This naturally causes chaos. The fix is
3835 not to close the socket in ip_connect(); the socket should be closed by
3836 the function that creates it. There was only one place in the code where
3837 this was missing, in the iplookup router, which I don't think anybody now
3838 uses, but I've fixed it anyway.
3840 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3841 well as to direct DNS lookups. Otherwise the handling of names in host
3842 lists is inconsistent and therefore confusing.
3848 PH/01 Two changes to the default runtime configuration:
3850 (1) Move the checks for relay_from_hosts and authenticated clients from
3851 after to before the (commented out) DNS black list checks.
3853 (2) Add control=submission to the relay_from_hosts and authenticated
3854 clients checks, on the grounds that messages accepted by these
3855 statements are most likely to be submissions.
3857 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3859 (1) Generate an error if the third argument for the ${prvs expansion is
3862 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3865 (3) Reset the variables that are obtained from the first argument of
3866 ${prvscheck and used in the second argument before leaving the code,
3867 because their memory is reclaimed, so using them afterwards may do
3870 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3871 one (it's much easier than Tom thought :-).
3873 (5) Because of (4), we can now allow for the use of $prvscheck_result
3874 inside the third argument.
3876 PH/03 For some reason, the default setting of PATH when running a command from
3877 a pipe transport was just "/usr/bin". I have changed it to
3880 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3881 anything to be listed in the output from -bV.
3883 PH/05 When a filter generated an autoreply, the entire To: header line was
3884 quoted in the delivery log line, like this:
3886 => >A.N.Other <ano@some.domain> <original@ddress> ...
3888 This has been changed so that it extracts the operative address. There
3889 may be more than one such address. If so, they are comma-separated, like
3892 => >ano@some.domain,ona@other.domain <original@ddress> ...
3894 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3895 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3896 not being looked up in the rDNS to get a host name, Exim was showing the
3897 IP address twice in Received: lines, even though the IP addresses were
3898 identical. For example:
3900 Received: from [1.2.3.4] (helo=[1.2.3.4])
3902 However, if the real host name was known, it was omitting the HELO data
3903 if it matched the actual IP address. This has been tidied up so that it
3904 doesn't show the same IP address twice.
3906 PH/07 When both +timestamp and +memory debugging was on, the value given by
3907 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3908 called by the debug printing, thereby overwriting the timestamp buffer.
3909 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3911 PH/08 When the original message was included in an autoreply transport, it
3912 always said "this is a copy of the message, including all the headers",
3913 even if body_only or headers_only was set. It now gives an appropriate
3916 PH/09 Applied a patch from the Sieve maintainer which:
3918 o fixes some comments
3919 o adds the (disabled) notify extension core
3920 o adds some debug output for the result of if/elsif tests
3921 o points to the current vacation draft in the documentation
3922 and documents the missing references header update
3926 o fixes a bug in processing the envelope test (when testing
3927 multiple envelope elements, the last element determined the
3930 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3931 Electronic Mail") by including:
3933 Auto-submitted: auto-generated
3935 in the messages that it generates (bounce messages and others, such as
3936 warnings). In the case of bounce messages for non-SMTP messages, there was
3937 also a typo: it was using "Auto_submitted" (underscore instead of
3938 hyphen). Since every message generated by Exim is necessarily in response
3939 to another message, thes have all been changed to:
3941 Auto-Submitted: auto-replied
3943 in accordance with these statements in the RFC:
3945 The auto-replied keyword:
3947 - SHOULD be used on messages sent in direct response to another
3948 message by an automatic process,
3950 - MUST NOT be used on manually-generated messages,
3952 - MAY be used on Delivery Status Notifications (DSNs) and Message
3953 Disposition Notifications (MDNs),
3955 - MUST NOT be used on messages generated by automatic or periodic
3956 processes, except for messages which are automatic responses to
3959 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3960 to the default Received: header definition.
3962 PH/12 Added log selector acl_warn_skipped (default on).
3964 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3965 variables because (a) they are in the wrong storage pool and (b) even if
3966 they were copied, it wouldn't work properly because of the caching.
3968 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3969 checking when decoding. Apparently there are clients that generate
3970 overlong encoded strings. Why am I not surprised?
3972 PH/15 If the first argument of "${if match_address" was not empty, but did not
3973 contain an "@" character, Exim crashed. Now it writes a panic log message
3974 and treats the condition as false.
3976 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3978 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3979 extension "envelope-auth". The code is finished and in agreement with
3980 other implementations, but there is no documentation so far and in fact,
3981 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3982 not changing the active code.
3984 Print executed "if" and "elsif" statements when debugging is used. This
3985 helps a great deal to understand what a filter does.
3987 Document more things not specified clearly in RFC3028. I had all this
3988 sorted out, when out of a sudden new issues came to my mind. Oops."
3990 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3993 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3994 canonical form (as documented). However, after a host name lookup from
3995 the IP address, check_host() was doing a simple string comparison with
3996 addresses acquired from the DNS when checking that the found name did
3997 have the original IP as one of its addresses. Since any found IPv6
3998 addresses are likely to be in abbreviated form, the comparison could
3999 fail. Luckily, there already exists a function for doing the comparison
4000 by converting both addresses to binary, so now that is used instead of
4001 the text comparison.
4003 PH/20 There was another similar case to PH/19, when a complete host name was
4004 given in a host list; looking up its IP address could give an abbreviated
4005 form, whereas the current host's name might or might not be abbreviated.
4006 The same fix has been applied.
4012 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4013 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4016 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4017 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4019 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4021 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4022 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4023 Also, for "Auto-Submitted", treat anything other than "no" as
4024 non-personal, in accordance with RFC 3834. (Previously it treated
4025 anything starting "auto-" as non-personal.)
4027 TF/01 The control=submission/name=... option had a problem with syntax
4028 errors if the name included a slash character. The /name= option
4029 now slurps the rest of the string, so it can include any characters
4030 but it must come last in the list of options (after /sender_retain
4033 PH/05 Some modifications to the interface to the fake nameserver for the new
4041 TK/01 Added the "success_on_redirect" address verification option. See
4042 NewStuff for rationale and an example.
4044 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4046 PH/02 Patch to exigrep to allow it to work on syslog lines.
4048 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4049 fread() to skip over the body file's header line, because in Cygwin the
4050 header line is locked and is inaccessible.
4052 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4053 co-exist for some time) to make it clear that it is the Exim ID that is
4054 referenced, not the Message-ID: header line.
4056 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4057 string_format() function, because snprintf() does not exist on all
4060 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4061 ${stat: expansion item.
4063 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4064 protocol synchronization error", to keep the pedants happy.
4066 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4067 well as for IRIX systems, when gcc is being used. See the host.c source
4070 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4072 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4075 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4076 transport and to the smtp transport in LMTP mode.
4078 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4080 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4081 run a filter in a subprocess. This could lead to confusion in subsequent
4082 lookups in the parent process. There should also be a search_tidyup() at
4083 the end of the subprocess.
4085 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4086 only if the host matched helo_try_verify_hosts, which caused the
4087 verification to occur when the EHLO/HELO command was issued. The ACL just
4088 tested the remembered result. Now, if a previous verification attempt has
4089 not happened, "verify = helo" does it there and then.
4091 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4093 TK/03 Fix log output including CR from clamd.
4095 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4096 exist provoked a memory error which could cause a segfault.
4098 PH/15 Installed PCRE 6.2
4100 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4102 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4103 of the problem. Specifically, suggested +O2 rather than +O1 for the
4106 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4108 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4109 record, the port was not correctly logged when the outgoing_port log
4110 selector was set (it logged the transort's default port).
4112 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4113 fallback_hosts, and "hosts" in the smtp transport.
4115 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4116 host errors such as "Connection refused".
4118 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4119 authentication with radiusclient 0.4.9:
4121 - Error returned from rc_read_config was caught wrongly
4122 - Username/password not passed on to radius server due to wrong length.
4124 The presumption is that some radiusclient API changes for 4.51/PH/17
4125 were not taken care of correctly. The code is still untested by me (my
4126 Linux distribution still has 0.3.2 of radiusclient), but it was
4127 contributed by a Radius user.
4129 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4130 expanding the "port" option of the smtp transport.
4132 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4133 while reading a MIME header. Thanks to Tom Hughes for a patch.
4135 PH/24 Include config.h inside local_scan.h so that configuration settings are
4138 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4139 that in an ACL for RCPT (for example), you can examine exactly what was
4142 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4143 commands, but it was not correctly comparing the address with the actual
4144 client host address. Thus, it would show the EHLO address in Received:
4145 header lines when this was not necessary.
4147 PH/27 Added the % operator to ${eval:}.
4149 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4150 it should be ignoring failures (because with -C, for example, it has lost
4151 privilege). It wasn't ignoring creation failures other than "already
4154 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4157 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4158 because an input line was too long, either on its own, or by virtue of
4159 too many continuations, the temporary file was not being removed, and the
4160 return code was incorrect.
4162 PH/31 Missing "BOOL" in function definition in filtertest.c.
4164 PH/32 Applied Sieve patches from the maintainer.
4166 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4168 PH/33 Added "verify = not_blind".
4170 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4171 Local/Makefile (with some defaults set). These are used in built scripts
4172 such as exicyclog, but they have never been used in the exim_install
4173 script (though there are many overriding facilities there). I have
4174 arranged that the exim_install script now takes note of these two
4177 PH/35 Installed configuration files for Dragonfly.
4179 PH/36 When a locally submitted message by a trusted user did not contain a
4180 From: header, and the sender address was obtained from -f or from an SMTP
4181 MAIL command, and the trusted user did not use -F to supply a sender
4182 name, $originator_name was incorrectly used when constructing a From:
4183 header. Furthermore, $originator_name was used for submission mode
4184 messages from external hosts without From: headers in a similar way,
4185 which is clearly wrong.
4187 PH/37 Added control=suppress_local_fixups.
4189 PH/38 When log_selector = +received_sender was set, and the addition of the
4190 sender made the log line's construction buffer exactly full, or one byte
4191 less than full, an overflow happened when the terminating "\n" was
4194 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4195 when the result of a list match is failure because a DNS lookup failed.
4197 PH/40 RM_COMMAND is now used in the building process.
4199 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4200 the "build-* directories that it finds.
4202 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4203 address in a domain literal was a prefix of an interface address.
4205 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4206 when verifying a sender address, unless rewrite_headers is false.
4208 PH/44 Wrote a long comment about why errors_to addresses are verified as
4209 recipients, not senders.
4211 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4212 the ratelimit ACL was added.
4214 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4216 PH/46 Added extra information about PostgreSQL errors to the error string.
4218 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4219 avoiding the need to install special zones in a real server. This is
4220 backwards compatible; if it can't find the fake resolver, it drops back.
4221 Thus, both old and new test suites can be run.
4223 TF/02 Added util/ratelimit.pl
4225 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4226 clock is set back in time.
4228 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4229 Candler <B.Candler@pobox.com>.
4231 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4232 OK for addresses that are the result of redirections.
4234 PH/48 A number of further additions for the benefit of the new test suite,
4235 including a fake gethostbyname() that interfaces to the fake DNS resolver
4238 TF/06 The fix for widen_domains has also been applied to qualify_single and
4239 search_parents which are the other dnslookup options that can cause
4242 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4245 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4247 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4248 service" if it encountered an MX record with an empty target hostname.
4249 The message is now "an MX or SRV record indicated no SMTP service".
4251 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4252 if the DNS of the sending site is misconfigured. This is quite a
4253 common situation. This change restores the behaviour of treating a
4254 helo verification defer as a failure.
4256 PH/51 If self=fail was set on a router, the bounce message did not include the
4257 actual error message.
4263 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4265 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4266 command itself ended in error, the underlying message about the transport
4267 filter timeout was being overwritten with the pipe command error. Now the
4268 underlying error message should be appended to the second error message.
4270 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4272 PH/02 Reduce the amount of output that "make" produces by default. Full output
4273 can still be requested.
4275 PH/03 The warning log line about a condition test deferring for a "warn" verb
4276 was being output only once per connection, rather than after each
4277 occurrence (because it was using the same function as for successful
4278 "warn" verbs). This seems wrong, so I have changed it.
4280 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4281 it should not have, which might have caused a crash in the right
4282 circumstances, but probably never did.
4284 PH/04 Installed a modified version of Tony Finch's patch to make submission
4285 mode fix the return path as well as the Sender: header line, and to
4286 add a /name= option so that you can make the user's friendly name appear
4289 TF/03 Added the control = fakedefer ACL modifier.
4291 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4292 Mark Lowes for thorough testing.
4294 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4296 TK/03 Merged latest SRS patch from Miles Wilton.
4298 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4299 with the definition in sysexits.h (which is #included earlier).
4300 Fortunately, Exim does not actually use EX_OK. The code used to try to
4301 preserve the sysexits.h value, by assuming that macro definitions were
4302 scanned for macro replacements. I have been disabused of this notion,
4303 so now the code just undefines EX_OK before #including unistd.h.
4305 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4306 in the smtp transport. When a block could not be written in a single
4307 write() function, the timeout was being re-applied to each part-write.
4308 This seems wrong - if the receiver was accepting one byte at a time it
4309 would take for ever. The timeout is now adjusted when this happens. It
4310 doesn't have to be particularly precise.
4312 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4313 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4315 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4316 without a domain if the check to <postmaster@domain> fails.
4318 SC/01 Eximstats: added -xls and the ability to specify output files
4319 (patch written by Frank Heydlauf).
4321 SC/02 Eximstats: use FileHandles for outputting results.
4323 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4325 SC/04 Eximstats: fixed display of large numbers with -nvr option
4327 SC/05 Eximstats: fixed merging of reports with empty tables.
4329 SC/06 Eximstats: added the -include_original_destination flag
4331 SC/07 Eximstats: removed tabs and trailing whitespace.
4333 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4335 TK/06 MBOX spool code: Add real "From " MBOX separator line
4336 so the .eml file is really in mbox format (even though
4337 most programs do not really care). Patch from Alex Miller.
4339 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4340 The latter is generated from $received_to and is only set if the
4341 message has one envelope recipient. SA can use these headers,
4342 obviously out-of-the-box. Patch from Alex Miller.
4344 PH/08 The ${def test on a variable was returning false if the variable's
4345 value was "0", contrary to what the specification has always said!
4346 The result should be true unless the variable is empty.
4348 PH/09 The syntax error of a character other than { following "${if
4349 def:variable_name" (after optional whitespace) was not being diagnosed.
4350 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4351 accidental colon was present, for example, could give incorrect results.
4353 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4354 result is used (not including appendfile, where other changes are about
4357 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4358 This involved changing a lot of size variables from int to off_t. It
4359 should work with maildirs and everything.
4361 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4362 spamd dying while we are connected to it.
4364 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4367 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4368 function for BDB 4.3.
4370 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4372 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4373 the sources for PCRE have been re-arranged and more files are now
4376 PH/15 The code I had for printing potentially long long variables in PH/11
4377 above was not the best (it lost precision). The length of off_t variables
4378 is now inspected at build time, and an appropriate printing format (%ld
4379 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4380 to be "long long int" or "long int". This is needed for the internal
4381 formatting function string_vformat().
4383 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4384 the configuration file to be ":syslog", then the script "guesses" where
4385 the logs files are, rather than using the compiled in default. In our
4386 case the guess is not the same as the compiled default, so the script
4387 suddenly stopped working when I started to use syslog. The patch checks
4388 to see if log_file_path is "". If so, it attempts to read it from exim
4389 with no configuration file to get the compiled in version, before it
4390 falls back to the previous guessing code."
4392 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4393 implementing BATV in an Exim configuration. See NewStuff for the gory
4396 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4397 Makefile that are specific to HP-UX.
4399 PH/18 If the "use_postmaster" option was set for a recipient callout together
4400 with the "random" option, the postmaster address was used as the MAIL
4401 FROM address for the random test, but not for the subsequent recipient
4402 test. It is now used for both.
4404 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4405 patch removes a few documentation additions to RFC 3028, because the
4406 latest draft now contains them. It adds the new en;ascii-case comparator
4407 and a new error check for 8bit text in MIME parts. Comparator and
4408 require names are now matched exactly. I enabled the subaddress
4409 extension, but it is not well tested yet (read: it works for me)."
4411 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4412 rework some of the code of TK/09 above to avoid the hardwired use of
4413 "%lld" and "long long". Replaced the call to snprintf() with a call to
4416 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4417 records point to non-existent hosts", "retry timeout exceeded", and
4418 "retry time not reached for any host after a long failure period".
4420 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4421 experimental DomainKeys support:
4423 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4424 (2) On an error such as an illegally used "control", the wrong name for
4425 the control was given.
4427 These problems did NOT occur unless DomainKeys support was compiled.
4429 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4431 PH/24 Added ${if match_ip condition.
4433 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4434 databases so that it will be absolutely obvious if a crash occurs in the
4435 DB library. This is a regular occurrence (often caused by mis-matched
4438 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4439 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4440 picked up on a user's system that detects such things. There doesn't seem
4441 to be a gcc warning option for this - only an attribute that has to be
4442 put on the function's prototype. It seems that in Fedora Core 4 they have
4443 set this on a number of new functions. No doubt there will be more in due
4446 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4447 specify a transport. However, if an address that was verified by such a
4448 router was the subject of a callout, Exim crashed because it tried to
4449 read the rcpt_include_affixes from the non-existent transport. Now it
4450 just assumes that the setting of that option is false. This bug was
4451 introduced by 4.51/PH/31.
4453 PH/28 Changed -d+all to exclude +memory, because that information is very
4454 rarely of interest, but it makes the output a lot bigger. People tend to
4455 do -d+all out of habit.
4457 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4458 code in os-type was giving problems when libc.so lives in lib64, like on
4461 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4462 aren't the modern standard, and it seems that some systems' include files
4463 don't always have them. Exim was already checking for some of the newer
4464 ones like T_AAAA, and defining it itself. I've added checks for all the
4465 record types that Exim uses.
4467 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4468 not automatically generating a new one, as it is supposed to. This
4469 prevented TLS from working. If the file did exist, but contained invalid
4470 data, a new version was generated, as expected. It was only the case of a
4471 non-existent file that was broken.
4473 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4474 with a change in libdomainkeys > 0.64.
4476 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4477 from DNS. If the selector record carries the flag, it now has
4478 precedence over the domain-wide flag.
4480 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4482 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4483 the use of a port name that isn't defined in /etc/services) occurred, the
4484 message was deferred as in a normal delivery, and thus remained on the
4485 spool, instead of being failed because of the mua_wrapper setting. This
4486 is now fixed, and I tidied up some of the mua_wrapper messages at the
4489 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4490 the messages in a hash of arrays rather than using individual hashes.
4491 This is a bit cleaner and results in dramatic memory savings, albeit
4492 at a slight CPU cost.
4494 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4495 as requested by Marc Sherman.
4497 SC/10 Eximstats: added histograms for user specified patterns as requested
4500 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4502 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4503 fopen() in the content-scanning modules that did not already have it.
4509 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4510 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4512 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4514 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4516 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4517 to test Sieve filters that use "vacation".
4519 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4520 that changes the way the GnuTLS parameters are stored in the cache file.
4521 The new format can be generated externally. For backward compatibility,
4522 if the data in the cache doesn't make sense, Exim assumes it has read an
4523 old-format file, and it generates new data and writes a new file. This
4524 means that you can't go back to an older release without removing the
4527 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4528 work if there are any delivery delays because "one_time" forces the
4529 parent to be marked "delivered", so its unseen clone is never tried
4530 again. For this reason, Exim now forbids the simultaneous setting of
4533 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4534 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4535 entry if you want to know the details. The fix, however, neglected to
4536 consider the case when local delivery batching is involved. The test for
4537 "previously delivered" was not happening when checking to see if an
4538 address could be batched with a previous (undelivered) one; under
4539 certain circumstances this could lead to multiple deliveries to the same
4542 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4543 in its include files, and this causes problems building Exim.
4545 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4546 header_syntax) but Exim was just ignoring anything given after a slash.
4547 In particular, this caused confusion with an attempt to use "verify =
4548 reverse_host_lookup/defer_ok". An error is now given when options are
4549 supplied for verify items that do not have them. (Maybe reverse_host_
4550 lookup should have a defer_ok option, but that's a different point.)
4552 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4553 defined by RFC 821) to 2048, because there were problems with some AUTH
4554 commands, and RFC 1869 says the size should be increased for extended
4555 SMTP commands that take arguments.
4557 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4560 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4561 "unknown" error; now it says that the functionality isn't in the binary.
4563 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4564 an address' error message when a string expansion fails (syntax or
4565 whatever). Otherwise the password may appear in the log. Following change
4566 PH/42 below, there is no longer a chance of it appearing in a bounce
4569 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4571 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4572 of its addresses was ever tried. (Bugzilla bug #2.)
4574 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4575 the result incorrectly in the debug output. (It correctly added a newline
4576 to what was transported.)
4578 TF/01 Added $received_time.
4580 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4581 commented out examples of how to interface to a virus scanner and to
4582 SpamAssassin. Also added commented examples of av_scanner and
4583 spamd_address settings.
4585 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4586 and controls are allowed in which ACLs. There were a couple of minor
4587 errors. Some of the entries in the conditions table (which is a table of
4588 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4589 negation of where the condition IS allowed.
4591 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4593 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4594 header file does not have a version number, so I've had to invent a new
4595 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4596 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4597 radiusclient), but it was contributed by a Radius user.
4599 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4600 files or directories, for OpenSSL.
4602 PH/19 When an Exim process that is running as root has to create an Exim log
4603 file, it does so in a subprocess that runs as exim:exim so as to get the
4604 ownership right at creation (otherwise, other Exim processes might see
4605 the file with the wrong ownership). There was no test for failure of this
4606 fork() call, which would lead to the process getting stuck as it waited
4607 for a non-existent subprocess. Forks do occasionally fail when resources
4608 run out. I reviewed all the other calls to fork(); they all seem to check
4611 PH/20 When checking for unexpected SMTP input at connect time (before writing
4612 the banner), Exim was not dealing correctly with a non-positive return
4613 from the read() function. If the client had disconnected by this time,
4614 the result was a log entry for a synchronization error with an empty
4615 string after "input=" when read() returned zero. If read() returned -1
4616 (an event I could not check), uninitialized data bytes were printed.
4617 There were reports of junk text (parts of files, etc) appearing after
4620 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4622 PH/22 Added support for macro redefinition, and (re)definition in between
4623 driver and ACL definitions.
4625 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4626 forgetting to use the resulting value; it was using the unexpanded value.
4628 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4629 hadn't been configured. The fix is from Juergen Kreileder, who
4630 understands it better than I do:
4632 "Here's what I see happening with three configured cyrus_sasl
4633 authenticators configured (plain, login, cram-md5):
4635 On startup auth_cyrus_sasl_init() gets called for each of these.
4636 This means three calls to sasl_listmech() without a specified mech_list.
4637 => SASL tests which mechs of all available mechs actually work
4638 => three warnings about OTP not working
4639 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4641 With the patch, sasl_listmech() also gets called three times. But now
4642 SASL's mech_list option is set to the server_mech specified in the the
4643 authenticator. Or in other words, the answer from sasl_listmech()
4644 gets limited to just the mech you're testing for (which is different
4646 => the return list contains just 'plain' or 'login', 'cram-md5' or
4647 nothing depending on the value of ob->server_mech.
4649 I've just tested the patch: Authentication still works fine,
4650 unavailable mechs specified in the exim configuration are still
4651 caught, and the auth.log warnings about OTP are gone."
4653 PH/25 When debugging is enabled, the contents of the command line are added
4654 to the debugging output, even when log_selector=+arguments is not
4657 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4658 answer is "GNU", and only if the return is "GNU/something" is the answer
4661 PH/27 $acl_verify_message is now set immediately after the failure of a
4662 verification in an ACL, and so is available in subsequent modifiers. In
4663 particular, the message can be preserved by coding like this:
4665 warn !verify = sender
4666 set acl_m0 = $acl_verify_message
4668 Previously, $acl_verify_message was set only while expanding "message"
4669 and "log_message" when a very denied access.
4671 PH/28 Modified OS/os.c-Linux with
4673 -#ifndef OS_LOAD_AVERAGE
4674 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4676 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4677 nomenclature these days.)
4679 PH/29 Installed patch from the Sieve maintainer that adds the options
4680 sieve_useraddress and sieve_subaddress to the redirect router.
4682 PH/30 In these circumstances:
4683 . Two addresses routed to the same list of hosts;
4684 . First host does not offer TLS;
4685 . First host accepts first address;
4686 . First host gives temporary error to second address;
4687 . Second host offers TLS and a TLS session is established;
4688 . Second host accepts second address.
4689 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4690 and peerdn, if requested) that were in fact used only for the second
4693 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4694 attention to any local part prefix or suffix that was matched by the
4695 router that accepted the address. It now behaves in the same way as it
4696 does for delivery: the affixes are removed from the local part unless
4697 rcpt_include_affixes is set on the transport.
4699 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4700 timeout during the DATA phase of an incoming message.
4702 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4703 applied a patch sanctioned by the Sieve maintainer.
4705 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4706 the uid or gid is negative. A case of a negative gid caused this to be
4707 noticed. The fix allows for either to be negative.
4709 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4710 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4713 PH/36 The change PH/12 above was broken. Fixed it.
4715 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4716 the grounds that routing the same address twice would always produce the
4717 same answer. This might have been true once, but it is certainly no
4718 longer true now. Routing a child address may depend on the previous
4719 routing that produced that child. Some complicated redirection strategies
4720 went wrong when messages had multiple recipients, and made Exim's
4721 behaviour dependent on the order in which the addresses were given.
4723 I have moved the duplicate checking until after the routing is complete.
4724 Exim scans the addresses that are assigned to local and remote
4725 transports, and removes any duplicates. This means that more work will be
4726 done, as duplicates will always all be routed, but duplicates are
4727 presumably rare, so I don't expect this is of any significance.
4729 For deliveries to pipes, files, and autoreplies, the duplicate checking
4730 still happens during the routing process, since they are not going to be
4733 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4734 It corrects a timeout issue with spamd. This is Ian's comment: "The
4735 background is that sometimes spamd either never reads data from a
4736 connection it has accepted, or it never writes response data. The exiscan
4737 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4738 blindly assumes that writes won't block so it may never time out."
4740 PH/39 Allow G after quota size as well as K and M.
4742 PH/40 The value set for $authenticated_id in an authenticator may not contain
4743 binary zeroes or newlines because the value is written to log lines and
4744 to spool files. There was no check on this. Now the value is run through
4745 the string_printing() function so that such characters are converted to
4746 printable escape sequences.
4748 PH/41 $message_linecount is a new variable that contains the total number of
4749 lines in the message. Compare $body_linecount, which is the count for the
4752 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4753 bounce and delay warning messages, except in certain special cases, which
4756 (a) An SMTP error message from a remote host;
4757 (b) A message specified in a :fail: redirection;
4758 (c) A message specified in a "fail" command in a system filter;
4759 (d) A message specified in a FAIL return from the queryprogram router;
4760 (e) A message specified by the cannot_route_message router option.
4762 In these cases only, Exim does include the error details in bounce and
4763 warning messages. There are also a few cases where bland messages such
4764 as "unrouteable address" or "local delivery error" are given.
4766 PH/43 $value is now also set for the "else" part of a ${run expansion.
4768 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4769 being worked on, but at least Exim now implements the latest version to
4772 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4773 process to complete was treated as a delivery failure, a timeout while
4774 writing the message to the pipe was logged, but erroneously treated as a
4775 successful delivery. Such timeouts include transport filter timeouts. For
4776 consistency with the overall process timeout, these timeouts are now
4777 treated as errors, giving rise to delivery failures by default. However,
4778 there is now a new Boolean option for the pipe transport called
4779 timeout_defer, which, if set TRUE, converts the failures into defers for
4780 both kinds of timeout. A transport filter timeout is now identified in
4783 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4784 systems where "make" and "gmake" are different, calling "gmake" at top
4785 level broke things. I've arranged for the value of $(MAKE) to be passed
4786 from the Makefile to this script so that it can call the same version of
4790 A note about Exim versions 4.44 and 4.50
4791 ----------------------------------------
4793 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4794 changes of various kinds. As a consequence, a big documentation update was
4795 needed. This delayed the release for rather longer than seemed good, especially
4796 in the light of a couple of (minor) security issues. Therefore, the changes
4797 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4798 release. So 4.44 and 4.50 are in effect two different branches that both start
4801 I have left the 4.50 change log unchanged; it contains all the changes since
4802 4.43. The change log for 4.44 is below; many of its items are identical to
4803 those for 4.50. This seems to be the most sensible way to preserve the
4804 historical information.
4810 1. Minor wording change to the doc/README.SIEVE file.
4812 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4813 computation of the current number of files was incorrect.
4815 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4816 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4819 4. Give more explanation in the error message when the command for a transport
4820 filter fails to execute.
4822 5. There are several places where Exim runs a non-Exim command in a
4823 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4824 was being done only for the command run by the queryprogram router. It is
4825 now done for all such subprocesses. The other cases are: ${run, transport
4826 filters, and the commands run by the lmtp and pipe transports.
4828 6. Added CONFIGURE_GROUP build-time option.
4830 7. Some older OS have a limit of 256 on the maximum number of file
4831 descriptors. Exim was using setrlimit() to set 1000 as a large value
4832 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4833 systems. I've change it so that if it can't get 1000, it tries for 256.
4835 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4836 was an oversight, and furthermore, ever since the addition of extra
4837 controls (e.g. 4.43/32), the checks on when to allow different forms of
4838 "control" were broken. There should now be diagnostics for all cases when a
4839 control that does not make sense is encountered.
4841 9. Added the /retain_sender option to "control=submission".
4843 10. $recipients is now available in the predata ACL (oversight).
4845 11. Tidy the search cache before the fork to do a delivery from a message
4846 received from the command line. Otherwise the child will trigger a lookup
4847 failure and thereby defer the delivery if it tries to use (for example) a
4848 cached ldap connection that the parent has called unbind on.
4850 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4851 of $address_data from the recipient verification was clobbered by the
4852 sender verification.
4854 13. The value of address_data from a sender verification is now available in
4855 $sender_address_data in subsequent conditions in the ACL statement.
4857 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4859 15. Added a new option "connect=<time>" to callout options, to set a different
4862 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4863 was its contents. (It was OK if the option was not defined at all.)
4865 17. A "Completed" log line is now written for messages that are removed from
4866 the spool by the -Mrm option.
4868 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4869 information about exactly what failed.
4871 19. Added -dd to debug only the daemon process.
4873 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4874 handles timeouts, both on the server side and network timeouts. Renamed the
4875 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4877 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4878 It is now set to "smtps".
4880 22. $host_address is now set to the target address during the checking of
4881 ignore_target_hosts.
4883 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4884 being passed; this would have caused $sender_host_name to have been used if
4885 matching the list had actually called for a host name (not very likely,
4886 since this list is usually IP addresses). A host name is now passed as
4889 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4890 code that specifies a non-restarting handler (typically sigaction() in
4891 modern systems) in an attempt to fix a rare and obscure crash bug.
4893 25. Narrowed the window for a race in the daemon that could cause it to ignore
4894 SIGCHLD signals. This is not a major problem, because they are used only to
4895 wake it up if nothing else does.
4897 26. A malformed maildirsize file could cause Exim to calculate negative values
4898 for the mailbox size or file count. Odd effects could occur as a result.
4899 The maildirsize information is now recalculated if the size or filecount
4902 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4903 support for a long time. Removed HAVE_SYS_VFS_H.
4905 28. Installed the latest version of exipick from John Jetmore.
4907 29. In an address list, if the pattern was not a regular expression, an empty
4908 subject address (from a bounce message) matched only if the pattern was an
4909 empty string. Non-empty patterns were not even tested. This was the wrong
4910 because it is perfectly reasonable to use an empty address as part of a
4911 database query. An empty address is now tested by patterns that are
4912 lookups. However, all the other forms of pattern expect the subject to
4913 contain a local part and a domain, and therefore, for them, an empty
4914 address still always fails if the pattern is not itself empty.
4916 30. Exim went into a mad DNS loop when attempting to do a callout where the
4917 host was specified on an smtp transport, and looking it up yielded more
4918 than one IP address.
4920 31. Re-factored the code for checking spool and log partition space into a
4921 function that finds that data and another that does the check. The former
4922 is then used to implement four new variables: $spool_space, $log_space,
4923 $spool_inodes, and $log_inodes.
4925 32. The RFC2047 encoding function was originally intended for short strings
4926 such as real names; it was not keeping to the 75-character limit for
4927 encoded words that the RFC imposes. It now respects the limit, and
4928 generates multiple encoded words if necessary. To be on the safe side, I
4929 have increased the buffer size for the ${rfc2047: expansion operator from
4932 33. It is now permitted to omit both strings after an "if" condition; if the
4933 condition is true, the result is "true". As before, when the second string
4934 is omitted, a false condition yields an empty string. This makes it less
4935 cumbersome to write custom ACL and router conditions.
4937 34. Failure to deliver a bounce message always caused it to be frozen, even if
4938 there was an errors_to setting on the router. The errors_to setting is now
4941 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4942 canonical form (fully expanded) before being placed in
4943 $sender_host_address.
4945 36. The table in the code that translates DNS record types into text (T_A to
4946 "A" for instance) was missing entries for NS and CNAME. It is just possible
4947 that this could have caused confusion if both these types were looked up
4948 for the same domain, because the text type is used as part of Exim's
4949 per-process caching. But the chance of anyone hitting this buglet seems
4952 37. The dnsdb lookup has been extended in a number of ways.
4954 (1) There is a new type, "zns", which walks up the domain tree until it
4955 finds some nameserver records. It should be used with care.
4957 (2) There is a new type, "mxh", which is like "mx" except that it returns
4958 just the host names, not the priorities.
4960 (3) It is now possible to give a list of domains (or IP addresses) to be
4961 looked up. The behaviour when one of the lookups defers can be
4962 controlled by a keyword.
4964 (4) It is now possible to specify the separator character for use when
4965 multiple records are returned.
4967 38. The dnslists ACL condition has been extended: it is now possible to supply
4968 a list of IP addresses and/or domains to be looked up in a particular DNS
4971 39. Added log_selector=+queue_time_overall.
4973 40. When running the queue in the test harness, wait just a tad after forking a
4974 delivery process, to get repeatability of debugging output.
4976 41. Include certificate and key file names in error message when GnuTLS fails
4977 to set them up, because the GnuTLS error message doesn't include the name
4978 of the failing file when there is a problem reading it.
4980 42. Allow both -bf and -bF in the same test run.
4982 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4984 44. The "Exiscan patch" is now merged into the mainline Exim source.
4986 45. Sometimes the final signoff response after QUIT could fail to get
4987 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4988 < 0 before doing a fflush(). This bug looks as though it goes back to the
4989 introduction of TLS in release 3.20, but "sometimes" must have been rare
4990 because the tests only now provoked it.
4992 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4993 (this can affect the format of dates).
4995 47. exim_tidydb, when checking for the continued existence of a message for
4996 which it has found a message-specific retry record, was not finding
4997 messages that were in split spool directories. Consequently, it was
4998 deleting retry records that should have stayed in existence.
5000 48. Steve fixed some bugs in eximstats.
5002 49. The SPA authentication driver was not abandoning authentication and moving
5003 on to the next authenticator when an expansion was forced to fail,
5004 contradicting the general specification for all authenticators. Instead it
5005 was generating a temporary error. It now behaves as specified.
5007 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5008 (the order specifies the preference for clients). The order is now AES256,
5009 AES128, 3DES, ARCFOUR128.
5011 51. Small patch to Sieve code - explicitly set From: when generating an
5014 52. Exim crashed if a remote delivery caused a very long error message to be
5015 recorded - for instance if somebody sent an entire SpamAssassin report back
5016 as a large number of 550 error lines. This bug was coincidentally fixed by
5017 increasing the size of one of Exim's internal buffers (big_buffer) that
5018 happened as part of the Exiscan merge. However, to be on the safe side, I
5019 have made the code more robust (and fixed the comments that describe what
5022 53. Now that there can be additional text after "Completed" in log lines (if
5023 the queue_time_overall log selector is set), a one-byte patch to exigrep
5024 was needed to allow it to recognize "Completed" as not the last thing in
5027 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5028 patch that reportedly fixes this has been added. I am not expert enough to
5029 create a test for it. This is what the patch creator wrote:
5031 "I found a little strange behaviour of ldap code when working with
5032 Windows 2003 AD Domain, where users was placed in more than one
5033 Organization Units. When I tried to give exim partial DN, the exit code
5034 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5035 But simultaneously result of request was absolutely normal ldap result,
5036 so I produce this patch..."
5038 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5039 REFERENCE, so I have modified the code to exclude the patch when that macro
5042 55. Some experimental protocols are using DNS PTR records for new purposes. The
5043 keys for these records are domain names, not reversed IP addresses. The
5044 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5045 leaves it alone. Component reversal etc. now happens only for IP addresses.
5048 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5050 57. Double the size of the debug message buffer (to 2048) so that more of very
5051 long debug lines gets shown.
5053 58. The exicyclog utility now does better if the number of log files to keep
5054 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5056 59. Two changes related to the smtp_active_hostname option:
5058 (1) $smtp_active_hostname is now available as a variable.
5059 (2) The default for smtp_banner uses $smtp_active_hostname instead
5060 of $primary_hostname.
5062 60. The host_aton() function is supposed to be passed a string that is known
5063 to be a valid IP address. However, in the case of IPv6 addresses, it was
5064 not checking this. This is a hostage to fortune. Exim now panics and dies
5065 if the condition is not met. A case was found where this could be provoked
5066 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5067 components; fortuitously, this particular loophole had already been fixed
5068 by change 4.50/55 above.
5070 If there are any other similar loopholes, the new check in host_aton()
5071 itself should stop them being exploited. The report I received stated that
5072 data on the command line could provoke the exploit when Exim was running as
5073 exim, but did not say which command line option was involved. All I could
5074 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5075 running as the user.
5078 61. There was a buffer overflow vulnerability in the SPA authentication code
5079 (which came originally from the Samba project). I have added a test to the
5080 spa_base64_to_bits() function which I hope fixes it.
5083 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5084 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5086 63. The daemon start-up calls getloadavg() while still root for those OS that
5087 need the first call to be done as root, but it missed one case: when
5088 deliver_queue_load_max is set with deliver_drop_privilege. This is
5089 necessary for the benefit of the queue runner, because there is no re-exec
5090 when deliver_drop_privilege is set.
5092 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5093 This has been fixed.
5095 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5096 was in use, was not putting the data itself into the right store pool;
5097 consequently, it could be overwritten for a subsequent message in the same
5098 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5101 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5103 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5104 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5105 the function were treating the return as a boolean value, which happened to
5106 work because 0=false and not-0=true, but is not correct code.
5108 68. The host_aton() function was not handling scoped IPv6 addresses (those
5109 with, for example, "%eth0" on the end) correctly.
5111 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5112 negated items (that is, ~something) in unsigned ints. Some compilers
5113 apparently mutter when there is no cast.
5115 70. If an address verification called from an ACL failed, and did not produce a
5116 user-specific message (i.e. there was only a "system" message), nothing was
5117 put in $acl_verify_message. In this situation, it now puts the system
5120 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5121 session; change 4.31/43 added the unwanted input to the log line - except
5122 that it did not do this in the start of session case. It now does.
5124 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5125 This is wrong and can cause the other end to generate a synchronization
5126 error if it is another Exim or anything else that does the synchronization
5127 check. A QUIT command is no longer sent after a timeout.
5129 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5130 during host lookups.
5132 74. The defer_ok option of callout verification was not working if it was used
5133 when verifying addresses in header lines, that is, for this case:
5135 verify = header_sender/callout=defer_ok
5137 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5138 those file descriptors could be used for SMTP connections. If anything
5139 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5140 could be sent to the SMTP client, causing chaos. The daemon now opens
5141 stdin, stdout, and stderr to /dev/null when it puts itself into the
5144 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5145 log by default. The user can override this with suitable Perl magic.
5147 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5148 the log message when discard triggers, was not working for the DATA ACL or
5149 for the non-SMTP ACL.
5151 78. Error message wording change in sieve.c.
5153 79. If smtp_accept_max_per_host was set, the number of connections could be
5154 restricted to fewer than expected, because the daemon was trying to set up
5155 a new connection before checking whether the processes handling previous
5156 connections had finished. The check for completed processes is now done
5157 earlier. On busy systems, this bug wouldn't be noticed because something
5158 else would have woken the daemon, and it would have reaped the completed
5161 80. If a message was submitted locally by a user whose login name contained one
5162 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5163 It caused a spool format error. I have fixed the spool reading code. A
5164 related problem was that the "from" clause in the Received: line became
5165 illegal because of the space(s). It is now covered by ${quote_local_part.
5167 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5170 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5172 83. Patch from Sieve maintainer to fix unterminated string problem in
5173 "vacation" handling.
5175 84. Some minor changes to the Linux configuration files to help with other
5176 OS variants using glibc.
5178 85. One more patch for Sieve to update vacation handling to latest spec.
5181 ----------------------------------------------------
5182 See the note above about the 4.44 and 4.50 releases.
5183 ----------------------------------------------------
5189 1. Change 4.43/35 introduced a bug that caused file counts to be
5190 incorrectly computed when quota_filecount was set in an appendfile
5193 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5194 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5197 3. Give more explanation in the error message when the command for a transport
5198 filter fails to execute.
5200 4. There are several places where Exim runs a non-Exim command in a
5201 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5202 was being done only for the command run by the queryprogram router. It is
5203 now done for all such subprocesses. The other cases are: ${run, transport
5204 filters, and the commands run by the lmtp and pipe transports.
5206 5. Some older OS have a limit of 256 on the maximum number of file
5207 descriptors. Exim was using setrlimit() to set 1000 as a large value
5208 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5209 systems. I've change it so that if it can't get 1000, it tries for 256.
5211 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5212 was an oversight, and furthermore, ever since the addition of extra
5213 controls (e.g. 4.43/32), the checks on when to allow different forms of
5214 "control" were broken. There should now be diagnostics for all cases when a
5215 control that does not make sense is encountered.
5217 7. $recipients is now available in the predata ACL (oversight).
5219 8. Tidy the search cache before the fork to do a delivery from a message
5220 received from the command line. Otherwise the child will trigger a lookup
5221 failure and thereby defer the delivery if it tries to use (for example) a
5222 cached ldap connection that the parent has called unbind on.
5224 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5225 of $address_data from the recipient verification was clobbered by the
5226 sender verification.
5228 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5229 was its contents. (It was OK if the option was not defined at all.)
5231 11. A "Completed" log line is now written for messages that are removed from
5232 the spool by the -Mrm option.
5234 12. $host_address is now set to the target address during the checking of
5235 ignore_target_hosts.
5237 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5238 being passed; this would have caused $sender_host_name to have been used if
5239 matching the list had actually called for a host name (not very likely,
5240 since this list is usually IP addresses). A host name is now passed as
5243 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5244 code that specifies a non-restarting handler (typically sigaction() in
5245 modern systems) in an attempt to fix a rare and obscure crash bug.
5247 15. Narrowed the window for a race in the daemon that could cause it to ignore
5248 SIGCHLD signals. This is not a major problem, because they are used only to
5249 wake it up if nothing else does.
5251 16. A malformed maildirsize file could cause Exim to calculate negative values
5252 for the mailbox size or file count. Odd effects could occur as a result.
5253 The maildirsize information is now recalculated if the size or filecount
5256 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5257 support for a long time. Removed HAVE_SYS_VFS_H.
5259 18. Updated exipick to current release from John Jetmore.
5261 19. Allow an empty sender to be matched against a lookup in an address list.
5262 Previously the only cases considered were a regular expression, or an
5265 20. Exim went into a mad DNS lookup loop when doing a callout where the
5266 host was specified on the transport, if the DNS lookup yielded more than
5269 21. The RFC2047 encoding function was originally intended for short strings
5270 such as real names; it was not keeping to the 75-character limit for
5271 encoded words that the RFC imposes. It now respects the limit, and
5272 generates multiple encoded words if necessary. To be on the safe side, I
5273 have increased the buffer size for the ${rfc2047: expansion operator from
5276 22. Failure to deliver a bounce message always caused it to be frozen, even if
5277 there was an errors_to setting on the router. The errors_to setting is now
5280 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5281 canonical form (fully expanded) before being placed in
5282 $sender_host_address.
5284 24. Updated eximstats to version 1.33
5286 25. Include certificate and key file names in error message when GnuTLS fails
5287 to set them up, because the GnuTLS error message doesn't include the name
5288 of the failing file when there is a problem reading it.
5290 26. Expand error message when OpenSSL has problems setting up cert/key files.
5293 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5294 (this can affect the format of dates).
5296 28. exim_tidydb, when checking for the continued existence of a message for
5297 which it has found a message-specific retry record, was not finding
5298 messages that were in split spool directories. Consequently, it was
5299 deleting retry records that should have stayed in existence.
5301 29. eximstats updated to version 1.35
5302 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5303 1.35 - bugfix such that pie charts by volume are generated correctly
5305 30. The SPA authentication driver was not abandoning authentication and moving
5306 on to the next authenticator when an expansion was forced to fail,
5307 contradicting the general specification for all authenticators. Instead it
5308 was generating a temporary error. It now behaves as specified.
5310 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5311 (the order specifies the preference for clients). The order is now AES256,
5312 AES128, 3DES, ARCFOUR128.
5314 31. Small patch to Sieve code - explicitly set From: when generating an
5317 32. Exim crashed if a remote delivery caused a very long error message to be
5318 recorded - for instance if somebody sent an entire SpamAssassin report back
5319 as a large number of 550 error lines. This bug was coincidentally fixed by
5320 increasing the size of one of Exim's internal buffers (big_buffer) that
5321 happened as part of the Exiscan merge. However, to be on the safe side, I
5322 have made the code more robust (and fixed the comments that describe what
5325 33. Some experimental protocols are using DNS PTR records for new purposes. The
5326 keys for these records are domain names, not reversed IP addresses. The
5327 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5328 leaves it alone. Component reversal etc. now happens only for IP addresses.
5331 34. The host_aton() function is supposed to be passed a string that is known
5332 to be a valid IP address. However, in the case of IPv6 addresses, it was
5333 not checking this. This is a hostage to fortune. Exim now panics and dies
5334 if the condition is not met. A case was found where this could be provoked
5335 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5336 components; fortuitously, this particular loophole had already been fixed
5337 by change 4.50/55 or 4.44/33 above.
5339 If there are any other similar loopholes, the new check in host_aton()
5340 itself should stop them being exploited. The report I received stated that
5341 data on the command line could provoke the exploit when Exim was running as
5342 exim, but did not say which command line option was involved. All I could
5343 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5344 running as the user.
5347 35. There was a buffer overflow vulnerability in the SPA authentication code
5348 (which came originally from the Samba project). I have added a test to the
5349 spa_base64_to_bits() function which I hope fixes it.
5352 36. The daemon start-up calls getloadavg() while still root for those OS that
5353 need the first call to be done as root, but it missed one case: when
5354 deliver_queue_load_max is set with deliver_drop_privilege. This is
5355 necessary for the benefit of the queue runner, because there is no re-exec
5356 when deliver_drop_privilege is set.
5358 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5359 was in use, was not putting the data itself into the right store pool;
5360 consequently, it could be overwritten for a subsequent message in the same
5361 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5364 38. Sometimes the final signoff response after QUIT could fail to get
5365 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5366 < 0 before doing a fflush(). This bug looks as though it goes back to the
5367 introduction of TLS in release 3.20, but "sometimes" must have been rare
5368 because the tests only now provoked it.
5374 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5375 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5376 Now it returns an integer. A number of calls were still expecting a T/F
5377 return. Fortuitously, in all cases, the tests worked in OK situations,
5378 which is the norm. However, things would have gone wrong on any write
5379 failures on the smtp file descriptor. This function is used when sending
5380 messages over SMTP and also when doing verify callouts.
5382 2. When Exim is called to do synchronous delivery of a locally submitted
5383 message (the -odf or -odi options), it no longer closes stderr before doing
5386 3. Implemented the mua_wrapper option.
5388 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5390 5. Implemented the functions header_remove(), header_testname(),
5391 header_add_at_position(), and receive_remove_recipient(), and exported them
5394 6. If an ACL "warn" statement specified the addition of headers, Exim already
5395 inserted X-ACL-Warn: at the start if there was no header name. However, it
5396 was not making this test for the second and subsequent header lines if
5397 there were newlines in the string. This meant that an invalid header could
5398 be inserted if Exim was badly configured.
5400 7. Allow an ACL "warn" statement to add header lines at the start or after all
5401 the Received: headers, as well as at the end.
5403 8. Added the rcpt_4xx retry error code.
5405 9. Added postmaster_mailfrom=xxx to callout verification option.
5407 10. Added mailfrom=xxxx to the callout verification option, for verify=
5410 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5411 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5413 12. Inserted some casts to stop certain compilers warning when using pointer
5414 differences as field lengths or precisions in printf-type calls (mostly
5415 affecting debugging statements).
5417 13. Added optional readline() support for -be (dynamically loaded).
5419 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5420 same clock tick as a message's arrival, so that its received time was the
5421 same as the "first fail" time on the retry record, and that message
5422 remained on the queue past the ultimate address timeout, every queue runner
5423 would try a delivery (because it was past the ultimate address timeout) but
5424 after another failure, the ultimate address timeout, which should have then
5425 bounced the address, did not kick in. This was a "< instead of <=" error;
5426 in most cases the first failure would have been in the next clock tick
5427 after the received time, and all would be well.
5429 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5430 being recognized when the domain list was tested by the match_domain
5431 condition in an expansion string.
5433 16. Added the ${str2b64: operator.
5435 17. Exim was always calling setrlimit() to set a large limit for the number of
5436 processes, without checking whether the existing limit was already
5437 adequate. (It did check for the limit on file descriptors.) Furthermore,
5438 errors from getrlimit() and setrlimit() were being ignored. Now they are
5439 logged to the main and panic logs, but Exim does carry on, to try to do its
5440 job under whatever limits there are.
5442 18. Imported PCRE 5.0.
5444 19. Trivial typo in log message " temporarily refused connection" (the leading
5447 20. If the log selector return_path_on_delivery was set and an address was
5448 redirected to /dev/null, the delivery process crashed because it assumed
5449 that a return path would always be set for a "successful" delivery. In this
5450 case, the whole delivery is bypassed as an optimization, and therefore no
5453 21. Internal re-arrangement: the function for sending a challenge and reading
5454 a response while authentication was assuming a zero-terminated challenge
5455 string. It's now changed to take a pointer and a length, to allow for
5456 binary data in such strings.
5458 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5460 23. Exim was not respecting finduser_retries when seeking the login of the
5461 uid under which it was called; it was always trying 10 times. (The default
5462 setting of finduser_retries is zero.) Also, it was sleeping after the final
5463 failure, which is pointless.
5465 24. Implemented tls_on_connect_ports.
5467 25. Implemented acl_smtp_predata.
5469 26. If the domain in control=submission is set empty, Exim assumes that the
5470 authenticated id is a complete email address when it generates From: or
5471 Sender: header lines.
5473 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5474 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5475 chown and chgrp in /bin and hostname in /usr/bin.
5477 28. Exim was keeping the "process log" file open after each use, just as it
5478 does for the main log. This opens the possibility of it remaining open for
5479 long periods when the USR1 signal hits a daemon. Occasional processlog
5480 errors were reported, that could have been caused by this. Anyway, it seems
5481 much more sensible not to leave this file open at all, so that is what now
5484 29. The long-running daemon process does not normally write to the log once it
5485 has entered its main loop, and it closes the log before doing so. This is
5486 so that log files can straightforwardly be renamed and moved. However,
5487 there are a couple of unusual error situations where the daemon does write
5488 log entries, and I had neglected to close the log afterwards.
5490 30. The text of an SMTP error response that was received during a remote
5491 delivery was being truncated at 512 bytes. This is too short for some of
5492 the long messages that one sometimes sees. I've increased the limit to
5495 31. It is now possible to make retry rules that apply only when a message has a
5496 specific sender, in particular, an empty sender.
5498 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5499 it possible to be selective about when SMTP synchronization is enforced.
5501 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5503 32. Implemented hosts_connection_nolog.
5505 33. Added an ACL for QUIT.
5507 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5510 35. Added mailbox_size and mailbox_filecount to appendfile.
5512 36. Added control = no_multiline_responses to ACLs.
5514 37. There was a bug in the logic of the code that waits for the clock to tick
5515 in the case where the clock went backwards by a substantial amount such
5516 that the microsecond fraction of "now" was more than the microsecond
5517 fraction of "then" (but the whole seconds number was less).
5519 38. Added support for the libradius Radius client library this is found on
5520 FreeBSD (previously only the radiusclient library was supported).
5526 1. When certain lookups returned multiple values in the form name=value, the
5527 quoting of the values was not always being done properly. Specifically:
5528 (a) If the value started with a double quote, but contained no whitespace,
5530 (b) If the value contained whitespace other than a space character (i.e.
5531 tabs or newlines or carriage returns) it was not quoted.
5532 This fix has been applied to the mysql and pgsql lookups by writing a
5533 separate quoting function and calling it from the lookup code. The fix
5534 should probably also be applied to nisplus, ibase and oracle lookups, but
5535 since I cannot test any of those, I have not disturbed their existing code.
5537 2. A hit in the callout cache for a specific address caused a log line with no
5538 reason for rejecting RCPT. Now it says "Previous (cached) callout
5539 verification failure".
5541 3. There was an off-by-one bug in the queryprogram router. An over-long
5542 return line was truncated at 256 instead of 255 characters, thereby
5543 overflowing its buffer with the terminating zero. As well as fixing this, I
5544 have increased the buffer size to 1024 (and made a note to document this).
5546 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5547 when Exim is waiting for an SMTP response from a remote server, Exim
5548 restarts its select() call on the socket, thereby resetting its timeout.
5549 This is not a problem when such interrupts are rare. Somebody set up a cron
5550 job to run exiwhat every 2 minutes, which is less than the normal select()
5551 timeout (5 or 10 minutes). This meant that the select() timeout never
5552 kicked in because it was always reset. I have fixed this by comparing the
5553 time when an interrupt arrives with the time at the start of the first call
5554 to select(). If more time than the timeout has elapsed, the interrupt is
5555 treated as a timeout.
5557 5. Some internal re-factoring in preparation for the addition of Sieve
5558 extensions (by MH). In particular, the "personal" test is moved to a
5559 separate function, and given an option for scanning Cc: and Bcc: (which is
5560 not set for Exim filters).
5562 6. When Exim created an email address using the login of the caller as the
5563 local part (e.g. when creating a From: or Sender: header line), it was not
5564 quoting the local part when it contained special characters such as @.
5566 7. Installed new OpenBSD configuration files.
5568 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5569 try to make them clearer.
5571 9. Callout options, other than the timeout value, were being ignored when
5572 verifying sender addresses in header lines. For example, when using
5574 verify = header_sender/callout=no_cache
5576 the cache was (incorrectly) being used.
5578 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5579 only the Cygwin environment.
5581 11. When return_path_on_delivery was set as a log selector, if different remote
5582 addresses in the same message used different return paths and parallel
5583 remote delivery occurred, the wrong values would sometimes be logged.
5584 (Whenever a remote delivery process finished, the return path value from
5585 the most recently started remote delivery process was logged.)
5587 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5588 lines when AUTH and/or TLS are in use. This is the "received protocol"
5589 field. Exim used to use "asmtp" for authenticated SMTP, without any
5590 indication (in the protocol name) for TLS use. Now it follows the RFC and
5591 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5592 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5593 names appear in log lines as well as in Received: header lines.
5595 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5596 extensions, and comparison tests, and to fix some bugs.
5598 14. Changes to the "personal" filter test:
5600 (1) The test was buggy in that it was just doing the equivalent of
5601 "contains" tests on header lines. For example, if a user's address was
5602 anne@some.where, the "personal" test would incorrectly be true for
5604 To: susanne@some.where
5606 This test is now done by extracting each address from the header in turn,
5607 and checking the entire address. Other tests that are part of "personal"
5608 are now done using regular expressions (for example, to check local parts
5609 of addresses in From: header lines).
5611 (2) The list of non-personal local parts in From: addresses has been
5612 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5613 taken from the Sieve specification recommendations.
5615 (3) If the message contains any header line starting with "List-" it is
5616 treated as non-personal.
5618 (4) The test for "circular" in the Subject: header line has been removed
5619 because it now seems ill-conceived.
5621 15. Minor typos in src/EDITME comments corrected.
5623 16. Installed latest exipick from John Jetmore.
5625 17. If headers_add on a router specified a text string that was too long for
5626 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5627 of string_sprintf() is now avoided.
5629 18. $message_body_size was not set (it was always zero) when running the DATA
5630 ACL and the local_scan() function.
5632 19. For the "mail" command in an Exim filter, no default was being set for
5633 the once_repeat time, causing a random time value to be used if "once" was
5634 specified. (If the value happened to be <= 0, no repeat happened.) The
5635 default is now 0s, meaning "never repeat". The "vacation" command was OK
5636 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5637 (I found it when inspecting the code).
5639 20. There is now an overall timeout for performing a callout verification. It
5640 defaults to 4 times the callout timeout, which applies to individual SMTP
5641 commands during the callout. The overall timeout applies when there is more
5642 than one host that can be tried. The timeout is checked before trying the
5643 next host. This prevents very long delays if there are a large number of
5644 hosts and all are timing out (e.g. when the network connections are timing
5645 out). The value of the overall timeout can be changed by specifying an
5646 additional sub-option for "callout", called "maxwait". For example:
5648 verify = sender/callout=5s,maxwait=20s
5650 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5651 to the end before writing, but this should make it even safer).
5653 22. Exim was forgetting that it had advertised PIPELINING for the second and
5654 subsequent messages on an SMTP connection. It was also not resetting its
5655 memory on STARTTLS and an internal HELO.
5657 23. When Exim logs an SMTP synchronization error within a session, it now
5658 records whether PIPELINING has been advertised or not.
5660 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5661 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5662 rather than long int.
5664 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5666 26. Added the never_mail option to autoreply.
5672 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5673 crash if the getsockname() call failed; this can happen if a connection is
5674 closed very soon after it is established. The problem was simply in the
5675 order in which certain operations were done, causing Exim to try to write
5676 to the SMTP stream before it had set up the file descriptor. The bug has
5677 been fixed by making things happen in the correct order.
5683 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5684 before the connection was closed, thus losing the rejection response.
5686 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5687 some early Solaris releases, but causes trouble in current releases where
5688 socklen_t is defined.
5690 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5693 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5696 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5697 that releasing the top of it at the end releases what was used for sub-
5698 expansions (unless the block got too big). However, discard this block if
5699 the first thing is a variable or header, so that we can use its block when
5700 it is dynamic (useful for very large $message_headers, for example).
5702 6. Lookups now cache *every* query, not just the most recent. A new, separate
5703 store pool is used for this. It can be recovered when all lookup caches are
5704 flushed. Lookups now release memory at the end of their result strings.
5705 This has involved some general refactoring of the lookup sources.
5707 7. Some code has been added to the store_xxx() functions to reduce the amount
5708 of flapping under certain conditions.
5710 8. log_incoming_interface used to affect only the <= reception log lines. Now
5711 it causes the local interface and port to be added to several more SMTP log
5712 lines, for example "SMTP connection from", and rejection lines.
5714 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5716 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5718 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5719 could be overwritten at the end of the current message (or the start of a
5720 new message if it was set in a HELO ACL). The value is now preserved for
5721 the duration of the SMTP connection.
5723 12. If a transport had a headers_rewrite setting, and a matching header line
5724 contained an unqualified address, that address was qualified, even if it
5725 did not match any rewriting rules. The underlying bug was that the values
5726 of the flags that permit the existence of unqualified sender and recipient
5727 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5728 non-local messages, and by -bnq for local messages) were not being
5729 preserved with the message after it was received.
5731 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5732 "next input=" as part of the text comprising the host identity instead of
5733 the correct text. The code was using the same buffer for two different
5734 strings. However, depending on which order the printing function evaluated
5735 its arguments, the bug did not always show up. Under Linux, for example, my
5736 test suite worked just fine.
5738 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5739 doesn't work with some older versions of Perl. It has been changed to "my",
5740 which in any case is probably the better facility to use.
5742 15. A really picky compiler found some instances of statements for creating
5743 error messages that either had too many or two few arguments for the format
5746 16. The size of the buffer for calls to the DNS resolver has been increased
5747 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5748 for addresses that have a lot of PTR records. This alleviates a problem; it
5749 does not fully solve it.
5751 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5752 buffer now truncates the list and logs the incident, which is the same
5753 action as happens when Exim is looking up a host name and its aliases.
5754 Previously in this situation something unpredictable would happen;
5755 sometimes it was "internal error: store_reset failed".
5757 18. If a server dropped the connection unexpectedly when an Exim client was
5758 using GnuTLS and trying to read a response, the client delivery process
5759 crashed while trying to generate an error log message.
5761 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5762 string, for example:
5764 warn message = H1: something\nH2: something
5766 the text was added as a single header line from Exim's point of view
5767 though it ended up OK in the delivered message. However, searching for the
5768 second and subsequent header lines using $h_h2: did not work. This has been
5769 fixed. Similarly, if a system filter added multiple headers in this way,
5770 the routers could not see them.
5772 20. Expanded the error message when iplsearch is called with an invalid key to
5773 suggest using net-iplsearch in a host list.
5775 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5776 ACLs are no longer actually imposed (and a message to that effect is
5779 22. If a "gecos" field in a passwd entry contained escaped characters, in
5780 particular, if it contained a \" sequence, Exim got it wrong when building
5781 a From: or a Sender: header from that name. A second bug also caused
5782 incorrect handling when an unquoted " was present following a character
5783 that needed quoting.
5785 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5786 was not being matched caselessly.
5788 24. Arranged for all hyphens in the exim.8 source to be escaped with
5791 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5792 properly. Recipient callout cache records were still being keyed to include
5793 the sender, even when use_sender was set false. This led to far more
5794 callouts that were necessary. The sender is no longer included in the key
5795 when use_sender is false.
5797 26. Added "control = submission" modifier to ACLs.
5799 27. Added the ${base62d: operator to decode base 62 numbers.
5801 28. dnsdb lookups can now access SRV records.
5803 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5804 the configuration file.
5806 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5807 (-v) mode. This makes the output for a verbose queue run more intelligible.
5809 31. Added a use_postmaster feature to recipient callouts.
5811 32. Added the $body_zerocount variable, containing the number of binary zero
5812 bytes in the message body.
5814 33. The time of last modification of the "new" subdirectory is now used as the
5815 "mailbox time last read" when there is a quota error for a maildir
5818 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5820 35. Added +ignore_unknown as a special item in host lists.
5822 36. Code for decoding IPv6 addresses in host lists is now included, even if
5823 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5824 address was recognized as an IP address, but was then not correctly decoded
5825 into binary, causing unexpected and incorrect effects when compared with
5832 1. Very minor rewording of debugging text in manualroute to say "list of
5833 hosts" instead of "hostlist".
5835 2. If verify=header_syntax was set, and a header line with an unqualified
5836 address (no domain) and a large number of spaces between the end of the
5837 name and the colon was received, the reception process suffered a buffer
5838 overflow, and (when I tested it) crashed. This was caused by some obsolete
5839 code that should have been removed. The fix is to remove it!
5841 3. When running in the test harness, delay a bit after writing a bounce
5842 message to get a bit more predictability in the log output.
5844 4. Added a call to search_tidyup() just before forking a reception process. In
5845 theory, someone could use a lookup in the expansion of smtp_accept_max_
5846 per_host which, without the tidyup, could leave open a database connection.
5848 5. Added the variables $recipient_data and $sender_data which get set from a
5849 lookup success in an ACL "recipients" or "senders" condition, or a router
5850 "senders" option, similar to $domain_data and $local_part_data.
5852 6. Moved the writing of debug_print from before to after the "senders" test
5855 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5856 problems for message scanning, either using a data ACL, or using
5857 local_scan() because the Received: header was not generated till after they
5858 were called (in order to set the time as the time of reception completion).
5859 I have revised the way this works. The header is now generated after the
5860 body is received, but before the ACL or local_scan() are called. After they
5861 are run, the timestamp in the header is updated.
5867 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5868 before starting a queue runner without re-exec. This happened only when
5869 deliver_drop_privilege was set or when the Exim user was set to root. The
5870 effect of the bug was that timeouts during subsequent deliveries caused
5871 crashes instead of being properly handled. The handler is now left at its
5872 default (and expected) setting.
5874 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5875 message, again when deliver_drop_privilege is set or Exim is run as root.
5876 The bug described in (1) was not present in this case, but the tidying up
5877 of the other signals was missing. I have made the two cases consistent.
5879 3. The ignore_target_hosts setting on a manualroute router was being ignored
5880 for hosts that were looked up using the /MX notation.
5882 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5885 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5886 operated on the sender address. After changing the $sender_address to <>
5887 for the sender address verify, Exim was re-instated it as the original
5888 (before rewriting) address, but remembering that it had rewritten it, so it
5889 wasn't rewriting it again. This bug also had the effect of breaking the
5890 sender address verification caching when the sender address was rewritten.
5892 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5893 This has been changed so that if the ip literal address matches
5894 ignore_target_hosts, the router declines.
5896 7. Added expansion conditions match_domain, match_address, and match_local_
5897 part (NOT match_host).
5899 8. The placeholder for the Received: header didn't have a length field set.
5901 9. Added code to Exim itself and to exim_lock to test for a specific race
5902 condition that could lead to file corruption when using MBX delivery. The
5903 issue is with the lockfile that is created in /tmp. If this file is removed
5904 after a process has opened it but before that process has acquired a lock,
5905 there is the potential for a second process to recreate the file and also
5906 acquire a lock. This could lead to two Exim processes writing to the file
5907 at the same time. The added code performs the same test as UW imapd; it
5908 checks after acquiring the lock that its file descriptor still refers to
5909 the same named file.
5911 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5912 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5915 11. Added the smtp_active_hostname option. If used, this will typically be made
5916 to depend on the incoming interface address. Because $interface_address is
5917 not set up until the daemon has forked a reception process, error responses
5918 that can happen earlier (such as "too many connections") no longer contain
5921 12. If an expansion in a condition on a "warn" statement fails because a lookup
5922 defers, the "warn" statement is abandoned, and the next ACL statement is
5923 processed. Previously this caused the whole ACL to be aborted.
5925 13. Added the iplsearch lookup type.
5927 14. Added ident_timeout as a log selector.
5929 15. Added tls_certificate_verified as a log selector.
5931 16. Added a global option tls_require_ciphers (compare the smtp transport
5932 option of the same name). This controls incoming TLS connections.
5934 17. I finally figured out how to make tls_require_ciphers do a similar thing
5935 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5936 before starting the TLS session.
5938 18. Tabs are now shown as \t in -bP output.
5940 19. If the log selector return_path_on_delivery was set, Exim crashed when
5941 bouncing a message because it had too many Received: header lines.
5943 20. If two routers both had headers_remove settings, and the first one included
5944 a superfluous trailing colon, the final name in the first list and the
5945 first name in the second list were incorrectly joined into one item (with a
5946 colon in the middle).
5952 1. Added -C and -D options to the exinext utility, mainly to make it easier
5953 to include in the automated testing, but these could be helpful when
5954 multiple configurations are in use.
5956 2. The exinext utility was not formatting the output nicely when there was
5957 an alternate port involved in the retry record key, nor when there was a
5958 message id as well (for retries that were specific to a specific message
5959 and a specific host). It was also confused by IPv6 addresses, because of
5960 the additional colons they contain. I have fixed the IPv4 problem, and
5961 patched it up to do a reasonable job for IPv6.
5963 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5964 delivery, the log line now contains "pipelined" if PIPELINING was used.
5966 4. An SMTP transport process used to panic and die if the bind() call to set
5967 an explicit outgoing interface failed. This has been changed; it is now
5968 treated in the same way as a connect() failure.
5970 5. A reference to $sender_host_name in the part of a conditional expansion
5971 that was being skipped was still causing a DNS lookup. This no longer
5974 6. The def: expansion condition was not recognizing references to header lines
5975 that used bh_ and bheader_.
5977 7. Added the _cache feature to named lists.
5979 8. The code for checking quota_filecount in the appendfile transport was
5980 allowing one more file than it should have been.
5982 9. For compatibility with Sendmail, the command line option
5990 and sets the incoming protocol and host name (for trusted callers). The
5991 host name and its colon can be omitted when only the protocol is to be set.
5992 Note the Exim already has two private options, -pd and -ps, that refer to
5993 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5994 "s", but I don't think that's a major issue.
5996 10. A number of refactoring changes to the code, none of which should affect
5999 (a) The number of logging options was getting close to filling up the
6000 32-bit word that was used as a bit map. I have split them into two classes:
6001 those that are passed in the argument to log_write(), and those that are
6002 only ever tested independently outside of that function. These are now in
6003 separate 32-bit words, so there is plenty of room for expansion again.
6004 There is no change in the user interface or the logging behaviour.
6006 (b) When building, for example, log lines, the code previously used a
6007 macro that called string_cat() twice, in order to add two strings. This is
6008 not really sufficiently general. Furthermore, there was one instance where
6009 it was actually wrong because one of the argument was used twice, and in
6010 one call a function was used. (As it happened, calling the function twice
6011 did not affect the overall behaviour.) The macro has been replaced by a
6012 function that can join an arbitrary number of extra strings onto a growing
6015 (c) The code for expansion conditions now uses a table and a binary chop
6016 instead of a serial search (which was left over from when there were very
6017 few conditions). Also, it now recognizes conditions like "pam" even when
6018 the relevant support is not compiled in: a suitably worded error message is
6019 given if an attempt is made to use such a condition.
6021 11. Added ${time_interval:xxxxx}.
6023 12. A bug was causing one of the ddress fields not to be passed back correctly
6024 from remote delivery subprocesses. The field in question was not being
6025 subsequently used, so this caused to problems in practice.
6027 13. Added new log selectors queue_time and deliver_time.
6029 14. Might have fixed a bug in maildirsizefile handling that threw up
6030 "unexpected character" debug warnings, and recalculated the data
6031 unnecessarily. In any case, I expanded the warning message to give more
6034 15. Added the message "Restricted characters in address" to the statements in
6035 the default ACL that block characters like @ and % in local parts.
6037 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6038 Three changes have been made:
6040 (a) There was a serious bug; a negative response to MAIL caused the whole
6041 recipient domain to be cached as invalid, thereby blocking all messages
6042 to all local parts at the same domain, from all senders. This bug has
6043 been fixed. The domain is no longer cached after a negative response to
6044 MAIL if the sender used is not empty.
6046 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6049 (c) A new callout option, "use_sender" has been added for people who want
6050 the modified behaviour.
6056 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6059 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6060 indeed breaks things for older releases.
6062 3. Added additional logging to the case where there is a problem reading data
6063 from a filter that is running in a subprocess using a pipe, in order to
6064 try to track down a specific problem.
6066 4. Testing facility fudge: when running in the test harness and attempting
6067 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6068 getting "No route to host". Convert this to a timeout.
6070 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6073 6. Some OS don't have socklen_t but use size_t instead. This affects the
6074 fifth argument of getsockopt() amongst other things. This is now
6075 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6076 can be set for individual OS. I have set it for SunOS5, OSF1, and
6077 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6078 some earlier ones do not.
6080 7. Change 4.30/15 was not doing the test caselessly.
6082 8. The standard form for an IPv6 address literal was being rejected by address
6083 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6084 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6085 this, as well as the form without the "IPv6" on the front (but only when
6086 address literals are enabled, of course).
6088 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6090 10. Exim crashed if a message with an empty sender address specified by -f
6091 encountered a router with an errors_to setting. This could be provoked only
6092 by a command such as
6096 where an empty string was supplied; "<>" did not hit this bug.
6098 11. Installed PCRE release 4.5.
6100 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6101 remained set. It is now erased.
6103 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6104 times from message ids (which are base 36 rather than the normal 62).
6106 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6107 were being counted as actual protocol errors, and logged if the log
6108 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6109 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6110 and DATA following a set of rejected RCPTs do not count as protocol errors.
6111 In other words, Exim assumes they were pipelined, though this may not
6112 actually be the case. Of course, in all cases the client gets an
6113 appropriate error code.
6115 15. If a lookup fails in an ACL condition, a message about the failure may
6116 be available; it is used if testing the ACL cannot continue, because most
6117 such messages specify what the cause of the deferral is. However, some
6118 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6119 that caused an old message to be retained and used if a later statement
6120 caused a defer, replacing the real cause of the deferral.
6122 16. If an IP address had so many PTR records that the DNS lookup buffer
6123 was not large enough to hold them, Exim could crash while trying to process
6124 the truncated data. It now detects and logs this case.
6126 17. Further to 4.21/58, another change has been made: if (and only if) the
6127 first line of a message (the first header line) ends with CRLF, a bare LF
6128 in a subsequent header line has a space inserted after it, so as not to
6129 terminate the header.
6131 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6132 unnecessarily, used atoi() instead of strtol(), and didn't check the
6133 termination when getting file sizes from file names by regex.
6135 19. Completely re-implemented the support for maildirsize files, in the light
6136 of a number of problems with the previous contributed implementation
6137 (4.30/29). In particular:
6139 . If the quota is zero, the maildirsize file is maintained, but no quota is
6142 . If the maildir directory does not exist, it is created before any attempt
6143 to write a maildirsize file.
6145 . The quota value in the file is just a cache; if the quota is changed in
6146 the transport, the new value overrides.
6148 . A regular expression is available for excluding directories from the
6151 20. The autoreply transport checks the characters in options that define the
6152 message's headers; it allows continued headers, but it was checking with
6153 isspace() after an embedded newline instead of explicitly looking for a
6156 21. If all the "regular" hosts to which an address was routed had passed their
6157 expiry times, and had not reached their retry times, the address was
6158 bounced, even if fallback hosts were defined. Now Exim should go on to try
6161 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6162 equivalent code in the SMTP transport. Some hosts send humungous responses
6163 to HELO/EHLO, more than 1024 it seems.
6165 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6166 gives compiler warnings in some environments. I've now done it "properly",
6169 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6170 (because of problems with the built-in one) was declared to return uschar *
6171 instead of char *, causing compiler failure.
6173 25. Fixed a file descriptor leak when processing alias/forward files.
6175 26. Fixed a minor format string issue in dbfn.c.
6177 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6179 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6180 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6183 29. When a sender address is verified, it is cached, to save repeating the test
6184 when there is more than one recipient in a message. However, when the
6185 verification involves a callout, it is possible for different callout
6186 options to be set for different recipients. It is too complicated to keep
6187 track of this in the cache, so now Exim always runs a verification when a
6188 callout is required, relying on the callout cache for the optimization.
6189 The overhead is duplication of the address routing, but this should not be
6192 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6193 to be verified with callout=postmaster, and the main callout worked but the
6194 postmaster check failed, the verification correctly failed. However, if a
6195 subsequent RCPT command asked for sender verification *without* the
6196 postmaster check, incorrect caching caused this verification also to fail,
6199 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6200 it was not caching the DNS options (qualify_single, search_parents) that
6201 were used when the lookup failed. A subsequent lookup with different
6202 options therefore always gave the same answer, though there were cases
6203 where it should not have. (Example: a "domains = !$mx_any" option on a
6204 dnslookup router: the "domains" option is always processed without any
6205 widening, but the router might have qualify_single set.) Now Exim uses the
6206 cached value only when the same options are set.
6208 32. Added John Jetmore's "exipick" utility to the distribution.
6210 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6211 than a timeout (e.g. a certificate is required, and is not provided), an
6212 Exim server now closes the connection immediately. Previously it waited for
6213 the client to close - but if the client is SSL, it seems that they each
6214 wait for each other, leading to a delay before one of them times out.
6216 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6217 maintained 0.8.x compatibility because I don't think many are using it, and
6218 it is clearly obsolete.
6220 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6223 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6224 client certificate was expired. A simple patch fixes this, though I don't
6225 understand the full logic of why the verify callback is called multiple
6228 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6229 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6230 which causes problems with some clients (such as the Certicom SSL Plus
6231 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6232 disables the coutermeasure allowing Eudora to connect."
6234 38. Exim was not checking that a write() to a log file succeeded. This could
6235 lead to Bad Things if a log got too big, in particular if it hit a file
6236 size limit. Exim now panics and dies if it cannot write to a log file, just
6237 as it does if it cannot open a log file.
6239 39. Modified OS/Makefile-Linux so that it now contains
6241 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6243 The two -D definitions ensure that Exim is compiled with large file
6244 support, which makes it possible to handle log files that are bigger than
6247 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6248 instance) a domain was checked against a named list that involved a lookup,
6249 causing $domain_data to be set, then another domain was checked against the
6250 same list, then the first domain was re-checked, the value of $domain_data
6251 after the final check could be wrong. In particular, if the second check
6252 failed, it could be set empty. This bug probably also applied to
6255 41. The strip_trailing_dot option was not being applied to the address given
6256 with the -f command-line option.
6258 42. The code for reading a message's header from the spool was incrementing
6259 $received_count, but never initializing it. This meant that the value was
6260 incorrect (doubled) while delivering a message in the same process in which
6261 it was received. In the most common configuration of Exim, this never
6262 happens - a fresh exec is done - but it can happen when
6263 deliver_drop_privilege is set.
6265 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6266 it now includes up to 150 characters of the unexpected data in the log
6269 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6270 and building data strings. The size of both of these buffers was 10 000
6271 bytes - far larger than anybody would *ever* want, thought I. Needless to
6272 say, somebody hit the limit. I have increased the maximum line length to
6273 20 000 and the maximum data length of concatenated lines to 100 000. I have
6274 also fixed two bugs, because there was no checking on these buffers. Tsk,
6275 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6276 buffer is too small.
6278 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6279 lsearch lookups. Now it does.
6281 46. When parsing a route_list item in a manualroute router, a fixed-length
6282 buffer was used for the list of hosts. I made this 1024 bytes long,
6283 thinking that nobody would ever have a list of hosts that long. Wrong.
6284 Somebody had a whole pile of complicated expansion conditions, and the
6285 string was silently truncated, leading to an expansion error. It turns out
6286 that it is easier to change to an unlimited length (owing to other changes
6287 that have happened since this code was originally written) than to build
6288 structure for giving a limitation error. The length of the item that
6289 expands into the list of hosts is now unlimited.
6291 47. The lsearch lookup could not handle data where the length of text line was
6292 more than 4095 characters. Such lines were truncated, leading to shortened
6293 data being returned. It should now handle lines of any length.
6295 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6296 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6299 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6300 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6301 no longer happens in comment lines. A list of replacements is now placed
6302 at the head of all of the source files, except those whose only change is
6303 to replace PERL_COMMAND in the very first #! line.
6305 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6306 messages on the queue, with a bottom-up merge sort, using code contributed
6307 by Michael Haardt. This should make operations like -bp somewhat faster on
6308 large queues. It won't affect queue runners, except when queue_run_in_order
6311 51. Installed eximstats 1.31 in the distribution.
6313 52. Added support for SRV lookups to the dnslookup router.
6315 53. If an ACL referred to $message_body or $message_body_end, the value was not
6316 reset for any messages that followed in the same SMTP session.
6318 54. The store-handling optimization for building very long strings was not
6319 differentiating between the different store pools. I don't think this
6320 actually made any difference in practice, but I've tidied it.
6322 55. While running the routers to verify a sender address, $sender_address
6323 was still set to the sender address. This is wrong, because when routing to
6324 send a bounce to the sender, it would be empty. Therefore, I have changed
6325 it so that, while verifying a sender address, $sender_address is set to <>.
6326 (There is no change to what happens when verifying a recipient address.)
6328 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6329 target A or AAAA records (if not already returned) without resetting the
6330 qualify_single or search_parents options of the DNS resolver. These are
6331 inappropriate in this case because the targets of MX and SRV records must
6332 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6333 target that, when qualified, matched something in the local domain. These
6334 two options are now turned off when doing these lookups.
6336 57. It seems that at least some releases of Reiserfs (which does not have the
6337 concept of a fixed number of inodes) returns zero and not -1 for the
6338 number of available inodes. This interacted badly with check_spool_inodes,
6339 which assumed that -1 was the "no such thing" setting. What I have done is
6340 to check that the total number of inodes is greater than zero before doing
6341 the test of how many are available.
6343 58. When a "warn" ACL statement has a log_message modifier, the message is
6344 remembered, and not repeated. This is to avoid a lot of repetition when a
6345 message has many recipients that cause the same warning to be written.
6346 However, Exim was preserving the list of already written lines for an
6347 entire SMTP session, which doesn't seem right. The memory is now reset if a
6348 new message is started.
6350 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6351 debugging output unless log_rewrite was also set.
6353 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6354 of (int)(handle) when we know that handle contains (void *)(-1).
6356 61. The Exim daemon panic-logs an error return when it closes the incoming
6357 connection. However "connection reset by peer" seems to be common, and
6358 isn't really an error worthy of noting specially, so that particular error
6361 62. When Exim is trying to find all the local interfaces, it used to panic and
6362 die if the ioctl to get the interface flags failed. However, it seems that
6363 on at least one OS (Solaris 9) it is possible to have an interface that is
6364 included in the list of interfaces, but for which you get a failure error
6365 for this call. This happens when the interface is not "plumbed" into a
6366 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6367 failure of the "get flags" call assumes that the interface is down.
6369 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6370 makes life easier for people who are doing arithmetic on fields extracted
6371 from dates, where you often get leading zeros that should not be
6372 interpreted as octal.
6374 64. Added qualify_domain to the redirect router, to override the global
6377 65. If a pathologically long header line contained very many addresses (the
6378 report of this problem mentioned 10 000) and each of them was rewritten,
6379 Exim could use up a very large amount of memory. (It kept on making new
6380 copies of the header line as it rewrote, and never released the old ones.)
6381 At the expense of a bit more processing, the header rewriting function has
6382 been changed so that it no longer eats memory in this way.
6384 66. The generation of the Received: header has been moved from the time that a
6385 message starts to be received, to the time that it finishes. The timestamp
6386 in the Received: header should now be very close to that of the <= log
6387 line. There are two side-effects of this change:
6389 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6390 logged header lines no longer include the local Received: line, because
6391 it has not yet been created. The same applies to a copy of the message
6392 that is returned to a non-SMTP sender when a message is rejected.
6394 (b) When a filter file is tested using -bf, no additional Received: header
6395 is added to the test message. After some thought, I decided that this
6398 This change does not affect the value of $received_for. It is still set
6399 after address rewriting, but before local_scan() is called.
6401 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6403 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6404 gave an unhelpful panic error message, and a defer error. I have managed to
6405 change this behaviour so that it now rejects any supplied certificate,
6406 which seems right, as the list of acceptable certificates is empty.
6408 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6409 gave an unhelpful defer error. I have not managed to make this reject any
6410 supplied certificates, but the error message it gives is "no certificate
6411 supplied", which is not helpful.
6413 70. exigrep's output now also includes lines that are not associated with any
6414 message, but which match the given pattern. Implemented by a patch from
6415 Martin Sluka, which also tidied up the Perl a bit.
6417 71. Recipient callout verification, like sender verification, was using <> in
6418 the MAIL FROM command. This isn't really the right thing, since the actual
6419 sender may affect whether the remote host accepts the recipient or not. I
6420 have changed it to use the actual sender in the callout; this means that
6421 the cache record is now keyed on a recipient/sender pair, not just the
6422 recipient address. There doesn't seem to be a real danger of callout loops,
6423 since a callout by the remote host to check the sender would use <>.
6424 [SEE ABOVE: changed after hitting problems.]
6426 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6427 temporary errors. However, in the case of such a code being given after
6428 the end of a data transmission (i.e. after ".") Exim was failing to write
6429 a retry record for the message. (Yes, there was some broken host that was
6430 actually sending 8xx at this point.)
6432 73. An unknown lookup type in a host list could cause Exim to panic-die when
6433 the list was checked. (An example that provoked this was putting <; in the
6434 middle of a list instead of at the start.) If this happened during a DATA
6435 ACL check, a -D file could be left lying around. This kind of configuration
6436 error no longer causes Exim to die; instead it causes a defer error. The
6437 incident is still logged to the main and panic logs.
6439 74. Buglet left over from Exim 3 conversion. The message "too many messages
6440 in one connection" was written to the rejectlog but not the mainlog, except
6441 when address rewriting (yes!) was being logged.
6443 75. Added write_rejectlog option.
6445 76. When a system filter was run not as root (that is, when system_filter_user
6446 was set), the values of the $n variables were not being returned to the
6447 main process; thus, they were not subsequently available in the $sn
6450 77. Added +return_path_on_delivery log selector.
6452 78. A connection timeout was being treated differently from recipients deferred
6453 when testing hosts_max_try with a message that was older than the host's
6454 retry timeout. (The host should not be counted, thus allowing all hosts to
6455 be tried at least once before bouncing.) This may have been the cause of an
6456 occasionally reported bug whereby a message would remain on the queue
6457 longer than the retry timeout, but would be bounced if a delivery was
6458 forced. I say "may" because I never totally pinned down the problem;
6459 setting up timeout/retry tests is difficult. See also the next item.
6461 79. The ultimate address timeout was not being applied to errors that involved
6462 a combination of host plus message (for example, a timeout on a MAIL
6463 command). When an address resolved to a number of possible hosts, and they
6464 were not all tried for each delivery (e.g. because of hosts_max_try), a
6465 message could remain on the queue longer than the retry timeout.
6467 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6470 81. Fixed an obscure SMTP outgoing bug which required at least the following
6471 conditions: (a) there was another message waiting for the same server;
6472 (b) the server returned 5xx to all RCPT commands in the first message so
6473 that the message was not completed; (c) the server dropped the connection
6474 or gave a negative response to the RSET that Exim sends to abort the
6475 transaction. The observed case was a dropped connection after DATA that had
6476 been sent in pipelining mode. That is, the server had advertised PIPELINING
6477 but was not implementing it correctly. The effect of the bug was incorrect
6478 behaviour, such as trying another host, and this could lead to a crash.
6484 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6485 and daemon.c were passed as pointers to ints; they should have been
6486 pointers to socklen_t variables (which are typically unsigned ints).
6488 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6491 3. Fixed a really odd bug that affected only the testing scheme; patching a
6492 certain fixed string in the binary changed the value of another string that
6493 happened to be identical to the end of the original first string.
6495 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6496 name", it returns that address as the IP address. On some operating
6497 systems (e.g. Solaris), it also passes back the IP address string as the
6498 "host name". However, on others (e.g. Linux), it passes back an empty
6499 string. Exim wasn't checking for this, and was changing the host name to an
6500 empty string, assuming it had been canonicalized.
6502 5. Although rare, it is permitted to have more than one PTR record for a given
6503 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6504 all the names associated with an address, because they do in Solaris.
6505 However, it seems that they do not in Linux for data that comes from the
6506 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6507 I found this out when I moved to a new Linux workstation and tried to run
6508 the Exim test suite.
6510 To get round this problem I have changed the code so that it now does its
6511 own call to the DNS to look up PTR records when searching for a host name.
6512 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6513 addresses that are only in /etc/hosts are still found.
6515 This behaviour is, however, controlled by an option called host_lookup_
6516 order, which defaults to "bydns:byaddr". If people want to use the other
6517 order, or indeed, just use one or the other means of lookup, they can
6518 specify it in this variable.
6520 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6521 some operating systems, this comes back from gethostbyaddr() as an empty
6522 string, and this is what Exim used to test for. However, it seems that in
6523 other systems, "." is yielded. Exim now tests for this case too.
6525 7. The values of check_spool_space and check_log_space are now held internally
6526 as a number of kilobytes instead of an absolute number of bytes. If a
6527 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6528 kilobyte. This means that much larger values can be stored.
6530 8. Exim monitor: an attempt to get the action menu when not actually pointing
6531 at a message produces an empty menu entitled "No message selected". This
6532 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6533 no entries in it ("Shell widget menu has zero width and/or height"). So I
6534 have added a single, blank menu entry in this case.
6536 9. Added ${quote_local_part.
6538 10. MIME decoding is now applied to the contents of Subject: header lines when
6541 11. Now that a reference to $sender_host_address automatically causes a reverse
6542 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6543 host lookup before query-style lookups in lists that might use this
6544 variable. This has therefore been abolished, and the "net-" prefix is no
6545 longer necessary for query-style lookups.
6547 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6548 have been a typo for LFLAGS, so it has been changed.
6550 13. The install script calls Exim with "-C /dev/null" in order to find the
6551 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6552 to be output. However, since Exim outputs its version number before the
6553 error, it didn't break the script. It just looked ugly. I fixed this by
6554 always allowing "-C /dev/null" if the caller is root.
6556 14. Ignore overlarge ACL variable number when reading spool file - insurance
6557 against a later release with more variables having written the file.
6559 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6560 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6561 this, as well as the form without the "IPv6" on the front.
6563 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6564 OS/Makefile-Darwin file.
6566 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6567 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6569 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6570 variable. However, some OS use ints for inodes. I've added cast to long int
6571 to get rid of the compiler warning.
6573 19. I had forgotten to lock out "/../" in configuration file names when
6574 ALT_CONFIG_PREFIX was set.
6576 20. Routers used for verification do not need to specify transports. However,
6577 if such a router generated a host list, and callout was configured, Exim
6578 crashed, because it could not find a port number from the (non-existent)
6579 transport. It now assumes port 25 in this circumstance.
6581 21. Added the -t option to exigrep.
6583 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6584 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6585 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6586 them is set, LOOKUP_LSEARCH is forced.
6588 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6590 24. Added sender and host information to the "rejected by local_scan()" log
6591 line; previously there was no indication of these.
6593 25. Added .include_if_exists.
6595 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6596 Linux. It turns out that not all file systems support this. Apparently some
6597 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6598 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6599 not supported on the file descriptor, is now ignored when Exim is trying to
6600 sync a directory. This applies only to Linux.
6602 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6604 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6605 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6608 29. Added support for maildirsize files from supplied patch (modified a bit).
6610 30. The use of :fail: followed by an empty string could lead Exim to respond to
6611 sender verification failures with (e.g.):
6613 550 Verification failed for <xxx>
6614 550 Sender verify failed
6616 where the first response line was missing the '-' that indicates it is not
6617 the final line of the response.
6619 31. The loop for finding the name of the user that called Exim had a hardwired
6620 limit of 10; it now uses the value of finduser_retries, which is used for
6621 all other user lookups.
6623 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6626 33. Exim was neglecting to zero errno before one call of strtol() when
6627 expanding a string and expecting an integer value. On some systems this
6628 resulted in spurious "integer overflow" errors. Also, it was casting the
6629 result into an int without checking.
6631 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6632 did not work. The code looks as if it has *never* worked, though it appears
6633 to have been documented since at least release 1.62. I have made it work.
6635 35. The "timeout_DNS" error in retry rules, also documented since at least
6636 1.62, also never worked. As it isn't clear exactly what this means, and
6637 clearly it isn't a major issue, I have abolished the feature by treating it
6638 as "timeout", and writing a warning to the main and panic logs.
6640 36. The display of retry rules for -brt wasn't always showing the error code
6643 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6644 timeout_connect_A, timeout_connect_MX.
6646 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6647 to the empty sender.
6649 39. The daemon was not analysing the content of -oX till after it had closed
6650 stderr and disconnected from the controlling terminal. This meant that any
6651 syntax errors were only noted on the panic log, and the return code from
6652 the command was 0. By re-arranging the code a little, I've made the
6653 decoding happen first, so such errors now appear on stderr, and the return
6654 code is 1. However, the actual setting up of the sockets still happens in
6655 the disconnected process, so errors there are still only recorded on the
6658 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6659 connections (as happens on some IP stacks) was logged at start up time as
6660 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6661 it from "IPv6 and IPv4", which means that two separate sockets are being
6664 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6665 says whether AF_INET or AF_INET6 was passed as an argument.
6667 42. Exiwhat output was messed up when time zones were included in log
6670 43. Exiwhat now gives more information about the daemon's listening ports,
6671 and whether -tls-on-connect was used.
6673 44. The "port" option of the smtp transport is now expanded.
6675 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6676 silently ignored. Now an error message is written to the main and panic
6679 46. There's a new ACL modifier called "logwrite" which writes to a log file
6680 as soon as it is encountered.
6682 47. Added $local_user_uid and $local_user_gid at routing time.
6684 48. Exim crashed when trying to verify a sender address that was being
6687 49. Exim was recognizing only a space character after ".include". It now also
6688 recognizes a tab character.
6690 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6691 extracting the relevant information from the specification. The man page no
6692 longer contains scrambled data for the -d option, and I've added a section
6693 at the front about calling Exim under different names.
6695 51. Added "extra_headers" argument to the "mail" command in filter files.
6697 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6700 53. Installed eximstats 1.29.
6702 54. Added transport_filter_timeout as a generic transport option.
6704 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6705 Cc: header lines. This was required by RFC 822, but it not required by RFC
6708 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6709 incoming messages that did not have them. Now it does so only if the
6710 message originates locally, that is, if there is no associated remote host
6711 address. When Resent- header lines are present, this applies to the Resent-
6712 lines rather than the non-Resent- lines.
6714 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6715 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6717 58. Messages for configuration errors now include the name of the main
6718 configuration file - useful now that there may be more than one file in a
6719 list (.included file names were always shown).
6721 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6722 for those rare installations that do not start the daemon as root or run it
6723 setuid root. I've cut out the call to initgroups() if the daemon is not
6726 60. The Exim user and group can now be bound into the binary as text strings
6727 that are looked up at the start of Exim's processing.
6729 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6731 62. Added $mailstore_basename variable.
6733 63. Installed patch to sieve.c from Michael Haardt.
6735 64. When Exim failed to open the panic log after failing to open the main log,
6736 the original message it was trying to log was written to stderr and debug
6737 output, but if they were not available (the usual case in production), it
6738 was lost. Now it is written to syslog before the two lines that record the
6739 failures to open the logs.
6741 65. Users' Exim filters run in subprocesses under the user's uid. It is
6742 possible for a "deliver" command or an alias in a "personal" command to
6743 provoke an address rewrite. If logging of address rewriting is configured,
6744 this fails because the process is not running as root or exim. There may be
6745 a better way of dealing with this, but for the moment (because 4.30 needs
6746 to be released), I have disabled address rewrite logging when running a
6747 filter in a non-root, non-exim process.
6753 1. The buildconfig auxiliary program wasn't quoting the value set for
6754 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6755 not defined. This bug was masked in 4.22 by the effect that was fixed in
6758 2. Some messages that were rejected after a message id was allocated were
6759 shown as "incomplete" by exigrep. It no longer does this for messages that
6760 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6762 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6763 have allow_domain_literals set, the ID did not get logged in the <= line.
6764 Domain literals are now always recognized in Message-ID: header lines.
6766 4. The first argument for a ${extract expansion item is the key name or field
6767 number. Leading and trailing spaces in this item were not being ignored,
6768 causing some misleading effects.
6770 5. When deliver_drop_privilege was set, single queue runner processes started
6771 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6772 same command in the process it spins off) were not dropping privilege.
6774 6. When the daemon running as "exim" started a queue runner, it always
6775 re-executed Exim in the spun-off process. This is a waste of effort when
6776 deliver_drop_privilege is set. The new process now just calls the
6777 queue-runner function directly.
6783 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6786 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6787 ignored. Though the use of -oP was forcing the writing of a pid file, it
6788 was always written to the default place.
6790 3. If the message "no IP address found for host xxxx" is generated during
6791 incoming verification, it is now followed by identification of the incoming
6792 connection (so you can more easily find what provoked it).
6794 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6796 5. Added some features to "harden" Exim a bit more against certain attacks:
6798 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6799 be put in Local/Makefile. This is like the never_users runtime option,
6800 but it cannot be overridden. The default setting is "root".
6802 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6803 prefix string with which any file named in a -C command line option
6806 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6807 is retained for -C and -D only if the caller of Exim is root. Without
6808 it, the exim user may also use -C and -D and retain privilege.
6810 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6811 command line option is disabled.
6813 6. Macro names set by the -D option must start with an upper case letter, just
6814 like macro names defined in the configuration file.
6816 7. Added "dereference=" facility to LDAP.
6818 8. Two instances of the typo "uknown" in the source files are fixed.
6820 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6821 the Configure-Makefile script screwed up while processing it.
6823 10. Incorporated PCRE 4.4.
6825 11. The SMTP synchronization check was not operating right at the start of an
6826 SMTP session. For example, it could not catch a HELO sent before the client
6827 waited for the greeting. There is now a check for outstanding input at the
6828 point when the greeting is written. Because of the duplex, asynchronous
6829 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6830 way, but not yet received, when the check is performed.
6832 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6833 on TCP/IP sockets, because this apparently causes some broken clients to
6836 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6837 unchanged) from the Cygwin maintainer.
6839 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6840 was supported instead of testing for mbx. Effectively a typo.
6842 15. The spa authenticator server code was not checking that the input it
6843 received was valid base64.
6845 16. The debug output line for the "set" modifier in ACLs was not showing the
6846 name of the variable that was being set.
6848 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6850 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6851 lookup on its own. Something else was needed to trigger the lookup. For
6852 example, a match in host_lookup or the need for a host name in a host list.
6853 Now, if $sender_host_name is referenced and the host name has not yet been
6854 looked up, a lookup is performed. If the lookup fails, the variable remains
6855 empty, and $host_lookup_failed is set to "1".
6857 19. Added "eqi" as a case-independent comparison operator.
6859 20. The saslauthd authentication condition could segfault if neither service
6860 nor realm was specified.
6862 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6863 error message that was logged was misleading, and incoming SMTP
6864 connections were dropped. The message is now more accurate, and temporary
6865 errors are given to SMTP connections.
6867 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6868 (except RSET) with a 421 error, until QUIT is received. However, it was
6869 failing to send a response to QUIT.
6871 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6872 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6873 ACL is now run after the helo_try_verify_hosts code.
6875 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6876 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6877 case-independent, so other case variants are also recognized). Apparently
6878 some systems use these upper case variants.
6880 25. If more than two messages were waiting for the same host, and a transport
6881 filter was specified for the transport, Exim sent two messages over the
6882 same TCP/IP connection, and then failed with "socket operation on non-
6883 socket" when it tried to send the third.
6885 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6887 27. The extern definition of crypt16() in expand.c was not being excluded when
6888 the OS had its own crypt16() function.
6890 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6891 as a preferred synonym for return_size_limit, both as an option and as an
6894 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6896 30. Changed the default configuration ACL to relax the local part checking rule
6897 for addresses that are not in any local domains. For these addresses,
6898 slashes and pipe symbols are allowed within local parts, but the sequence
6899 /../ is explicitly forbidden.
6901 31. SPA server authentication was not clearing the challenge buffer before
6904 32. log_message in a "warn" ACL statement was writing to the reject log as
6905 well as to the main log, which contradicts the documentation and doesn't
6906 seem right (because no rejection is happening). So I have stopped it.
6908 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6909 However, I am unable to do any testing of this.
6911 34. Fixed an infelicity in the appendfile transport. When checking directories
6912 for a mailbox, to see if any needed to be created, it was accidentally
6913 using path names with one or more superfluous leading slashes; tracing
6914 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6916 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6917 contents are added to the log line that is written for every discarded
6918 recipient. (Previously a log_message setting was ignored.)
6920 36. The ${quote: operator now quotes the string if it is empty.
6922 37. The install script runs exim in order to find its version number. If for
6923 some reason other than non-existence or emptiness, which it checks, it
6924 could not run './exim', it was installing it with an empty version number,
6925 i.e. as "exim-". This error state is now caught, and the installation is
6928 38. An argument was missing from the function that creates an error message
6929 when Exim fails to connect to the socket for saslauthd authentication.
6930 This could cause Exim to crash, or give a corrupted message.
6932 39. Added isip, isip4, and isip6 to ${if conditions.
6934 40. The ACL variables $acl_xx are now saved with the message, and can be
6935 accessed later in routers, transports, and filters.
6937 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6938 strings in the file are not string-expanded.
6940 42. If a MAIL command specified a SIZE value that was too large to fit into an
6941 int variable, the check against message_size_limit failed. Such values are
6942 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6943 day this will have to be increased, but I don't think I want to be around
6944 when emails are that large.
6951 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6952 iconv() is not standard in FreeBSD.
6954 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6955 IPv6 enabled. The observed symptom was a segmentation fault on return from
6956 the function os_common_find_running_interfaces() in src/os.c.
6958 3. In the check_special_case() function in daemon.c I had used "errno" as an
6959 argument name, which causes warnings on some systems. This was basically a
6960 typo, since it was named "eno" in the comments!
6962 4. The code that waits for the clock to tick (at a resolution of some fraction
6963 of a second) so as to ensure message-id uniqueness was always waiting for
6964 at least one whole tick, when it could have waited for less. [This is
6965 almost certainly not relevant at current processor speeds, where it is
6966 unlikely to ever wait at all. But we try to future-proof.]
6968 5. The function that sleeps for a time interval that includes fractions of a
6969 second contained a race. It did not block SIGALRM between setting the
6970 timer, and suspending (a couple of lines later). If the interval was short
6971 and the sigsuspend() was delayed until after it had expired, the suspension
6972 never ended. On busy systems this could lead to processes getting stuck for
6975 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6976 process, before it forks any delivery processes. The open lookup caching
6977 mechanism meant that the open file or database connection was passed into
6978 the delivery process. The problem was that delivery processes always tidy
6979 up cached lookup data. This could cause a problem for the next delivery
6980 process started by the queue runner, because the external queue runner
6981 process does not know about the closure. So the next delivery process
6982 still has data in the lookup cache. In the case of a file lookup, there was
6983 no problem because closing a file descriptor in a subprocess doesn't affect
6984 the parent. However, if the lookup was caching a connection to a database,
6985 the connection was closed, and the second delivery process was likely to
6986 see errors such as "PGSQL: query failed: server closed the connection
6987 unexpectedly". The problem has been fixed by closing all cached lookups
6988 in a queue runner before running a delivery process.
6990 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6991 seem to have the "const" qualifier which it has on other OS. I've
6994 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6995 *of the same type* (client or server) with the same public name that an
6996 error should be diagnosed.
6998 9. When Exim looked up a host name for an IP address, but failed to find the
6999 original IP address when looking up the host name (a safety check), it
7000 output the message "<ip address> does not match any IP for NULL", which was
7001 confusing, to say the least. The bug was that the host name should have
7002 appeared instead of "NULL".
7004 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7005 user that is built into the binary, and the -C or -D options is used, root
7006 privilege is dropped before the configuration file is read. In addition,
7007 logging is switched to stderr instead of the normal log files. If the
7008 configuration then re-defines the Exim user, the unprivileged environment
7009 is probably not what is expected, so Exim logs a panic warning message (but
7012 However, if deliver_drop_privilege is set, the unprivileged state may well
7013 be exactly what is intended, so the warning has been cut out in that case,
7014 and Exim is allowed to try to write to its normal log files.
7020 1. smtp_return_error_details was not giving details for temporary sender
7021 or receiver verification errors.
7023 2. Diagnose a configuration error if two authenticators have the same public
7026 3. Exim used not to create the message log file for a message until the first
7027 delivery attempt. This could be confusing when incoming messages were held
7028 for policy or load reasons. The message log file is now created at the time
7029 the message is received, and an initial "Received" line is written to it.
7031 4. The automatically generated man page for command line options had a minor
7032 bug that caused no ill effects; however, a more serious problem was that
7033 the procedure for building the man page automatically didn't always
7034 operate. Consequently, release 4.20 contains an out-of-date version. This
7035 shouldn't happen again.
7037 5. When building Exim with embedded Perl support, the script that builds the
7038 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7039 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7041 6. The freeze_tell option was not being used for messages that were frozen on
7042 arrival, either by an ACL or by local_scan().
7044 7. Added the smtp_incomplete_transaction log selector.
7046 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7047 was accepting AUTH without a new EHLO.
7049 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7050 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7053 10. Logging of TCP/IP connections (when configured) now happens in the main
7054 daemon process instead of the child process, so that the TCP/IP connection
7055 count is more accurate (but it can never be perfect).
7057 11. The use of "drop" in a nested ACL was not being handled correctly in the
7058 outer ACL. Now, if condition failure induced by the nested "drop" causes
7059 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7060 or "require"), the connection is dropped.
7062 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7063 that yield "discard" can now be used with an "accept" or a "discard" verb,
7064 but an error is generated for any others (because I can't see a useful way
7065 to define what should happen).
7067 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7068 are now processed in the same way as lines in the Exim configuration file.
7069 In particular, continuation lines are supported.
7071 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7073 15. Added -ti meaning -t -i.
7075 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7076 domains, and warn by logging if others are found.
7078 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7079 returned from GIFCONF when Exim is trying to find the list of interfaces on
7080 a host. The code in os.c has been modified to copy each ifreq to an aligned
7081 structure in all cases.
7083 Also, in some cases, the returned ifreq's were being copied to a 'struct
7084 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7085 means the last couple of bytes of an IPv6 address could be chopped if the
7086 ifreq contained only a normal sockaddr (14 bytes storage).
7088 18. Named domain lists were not supported in the hosts_treat_as_local option.
7089 An entry such as +xxxx was not recognized, and was treated as a literal
7092 19. Ensure that header lines added by a DATA ACL are included in the reject log
7093 if the ACL subsequently rejects the message.
7095 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7096 MD5 (which is deprecated).
7098 21. When testing a filter file using -bf, Exim was writing a message when it
7099 took the sender from a "From " line in the message, but it was not doing so
7100 when it took $return_path from a Return-Path: header line. It now does.
7102 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7103 with a valid header line field name (a series of printing characters
7104 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7106 23. Changed "disc" in the source to "disk" to conform to the documentation and
7107 the book and for uniformity.
7109 24. Ignore Sendmail's -Ooption=value command line item.
7111 25. When execve() failed while trying to run a command in a pipe transport,
7112 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7113 could be confused with a return value of 69 from the command itself. This
7114 has been changed to 127, the value the shell returns if it is asked to run
7115 a non-existent command. The wording for the related log line suggests a
7116 non-existent command as the problem.
7118 26. If received_header_text expands to an empty string, do not add a Received:
7119 header line to the message. (Well, it adds a token one on the spool, but
7120 marks it "old" so that it doesn't get used or transmitted.)
7122 27. Installed eximstats 1.28 (addition of -nt option).
7124 28. There was no check for failure on the call to getsockname() in the daemon
7125 code. This can fail if there is a shortage of resources on the system, with
7126 ENOMEM, for example. A temporary error is now given on failure.
7128 29. Contrary to the C standard, it seems that in some environments, the
7129 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7130 program. Exim now does this explicitly; it affects the formatting of
7131 timestamps using strftime().
7133 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7134 complaints. I've now initialized all the variables, to avoid this.
7136 32. Header lines added by a system filter were not being "seen" during
7137 transport-time rewrites.
7139 33. The info_callback() function passed to OpenSSL is set up with type void
7140 (*)(SSL *, int, int), as described somewhere. However, when calling the
7141 function (actually a macro) that sets it up, the type void(*)() is
7142 expected. I've put in a cast to prevent warnings from picky compilers.
7144 34. If a DNS black list lookup found a CNAME record, but there were no A
7145 records associated with the domain it pointed at, Exim crashed.
7147 35. If a DNS black list lookup returned more than one A record, Exim ignored
7148 all but the first. It now scans all returned addresses if a particular IP
7149 value is being sought. In this situation, the contents of the
7150 $dnslist_value variable are a list of all the addresses, separated by a
7153 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7154 to accept a host name and all its aliases if the forward lookup for any of
7155 them yielded the IP address of the incoming connection. Now it accepts only
7156 those names whose forward lookup yields the correct IP address. Any other
7157 names are discarded. This closes a loophole whereby a rogue DNS
7158 administrator could create reverse DNS records to break through a
7159 wildcarded host restriction in an ACL.
7161 37. If a user filter or a system filter that ran in a subprocess used any of
7162 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7163 the wrong values were passed to the pipe command ($thisaddress had the
7164 value of $0, $0 had the value of $1, etc). This bug was introduced by
7165 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7167 38. Improved the line breaking for long SMTP error messages from ACLs.
7168 Previously, if there was no break point between 40 and 75 characters, Exim
7169 left the rest of the message alone. Two changes have been made: (a) I've
7170 reduced the minimum length to 35 characters; (b) if it can't find a break
7171 point between 35 and 75 characters, it looks ahead and uses the first one
7172 that it finds. This may give the occasional overlong line, but at least the
7173 remaining text gets split now.
7175 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7176 file descriptors might be low, and that setting 1000 would always raise it.
7177 It turns out that in some environments, the limit is already over 1000 and
7178 that lowering it causes trouble. So now Exim takes care not to decrease it.
7180 40. When delivering a message, the value of $return_path is set to $sender_
7181 address at the start of routing (routers may change the value). By an
7182 oversight, this default was not being set up when an address was tested by
7183 -bt or -bv, which affected the outcome if any router or filter referred to
7186 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7187 log only when "message" or "log_message" are set. However, if one of the
7188 conditions was an address verification, or a call to a nested ACL, the
7189 messages generated by the underlying test were being passed through. This
7190 no longer happens. The underlying message is available in $acl_verify_
7191 message for both "message" and "log_message" expansions, so it can be
7192 passed through if needed.
7194 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7195 new expansion $bh_ to give the encoded byte string without charset
7196 translation. Translation happens only if iconv() is available; HAVE_ICONV
7197 indicates this at build time. HEADERS_CHARSET gives the charset to
7198 translate to; headers_charset can change it in the configuration, and
7199 "headers charset" can change it in an individual filter file.
7201 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7202 that creates RFC 2047 encoded "words" labels them as that charset instead
7203 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7204 expansion operator; (ii) when Exim creates a From: line for a local
7205 message; (iii) when a header line is rewritten to include a "phrase" part.
7207 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7208 buggy, causing it to skip the first lines of messages whose message ID
7209 ended in 'D'. This would not have bitten before Exim release 4.14, because
7210 message IDs were unlikely to end in 'D' before then. The effect was to have
7211 incorrect size information for certain domains.
7213 45. #include "config.h" was missing at the start of the crypt16.c module. This
7214 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7217 46. If there was a timeout during a "random" callout check, Exim treated it as
7218 a failure of the random address, and carried on sending RSET and the real
7219 address. If the delay was just some slowness somewhere, the response to the
7220 original RCPT would be taken as a response to RSET and so on, causing
7221 mayhem of various kinds.
7223 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7224 when I implemented it. It didn't allow for the fact that some option values
7225 may legitimately be negative (e.g. size_addition), and it didn't even do
7226 the right test for positive values.
7228 48. Domain names in DNS records are case-independent. Exim always looks them up
7229 in lower case. Some resolvers return domain names in exactly the case they
7230 appear in the zone file, that is, they may contain uppercase letters. Not
7231 all resolvers do this - some return always lower case. Exim was treating a
7232 change of case by a resolver as a change of domain, similar to a widening
7233 of a domain abbreviation. This triggered its re-routing code and so it was
7234 trying to route what was effectively the same domain again. This normally
7235 caused routing to fail (because the router wouldn't handle the domain
7236 twice). Now Exim checks for this case specially, and just changes the
7237 casing of the domain that it ultimately uses when it transmits the message
7240 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7243 50. If a filter generated a file delivery with a non-absolute name (possible if
7244 no home directory exists for the router), the forbid_file option was not
7247 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7248 the existing equality matching.
7250 52. Exim was using ints instead of ino_t variables in some places where it was
7251 dealing with inode numbers.
7253 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7254 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7255 called TMPDIR, and if it finds it is different, it changes its value.
7257 54. The smtp_printf() function is now made available to local_scan() so
7258 additional output lines can be written before returning. There is also an
7259 smtp_fflush() function to enable the detection of a dropped connection.
7260 The variables smtp_input and smtp_batched_input are exported to
7263 55. Changed the default runtime configuration: the message "Unknown user"
7264 has been removed from the ACL, and instead placed on the localuser router,
7265 using the cannot_route_message feature. This means that any verification
7266 failures that generate their own messages won't get overridden. Similarly,
7267 the "Unrouteable address" message that was in the ACL for unverifiable
7268 relay addresses has also been removed.
7270 56. Added hosts_avoid_esmtp to the smtp transport.
7272 57. The exicyclog script was not checking for the esoteric option
7273 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7274 will work only if exicyclog is run under the appropriate euid.
7276 58. Following a discussion on the list, the rules by which Exim recognises line
7277 endings on incoming messages have been changed. The -dropcr and drop_cr
7278 options are now no-ops, retained only for backwards compatibility. The
7279 following line terminators are recognized: LF CRLF CR. However, special
7280 processing applies to CR:
7282 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7283 nor a local message in the state where . is a terminator.
7285 (ii) If a bare CR is encountered in a header line, an extra space is added
7286 after the line terminator so as not to end the header. The reasoning
7287 behind this is that bare CRs in header lines are most likely either
7288 to be mistakes, or people trying to play silly games.
7290 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7291 was being incorrectly given as 18 bytes larger than it should have been.
7292 This is a VOB (very old bug).
7294 60. This may never have affected anything current, but just in case it has:
7295 When the local host is found other than at the start of a list of hosts,
7296 the local host, those with the same MX, and any that follow, are discarded.
7297 When the list in question was part of a longer list of hosts, the following
7298 hosts (not currently being processed) were also being discarded. This no
7299 longer happens. I'm not sure if this situation could ever has previously
7302 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7305 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7306 header. This is something that is recommended in a new Internet Draft, and
7307 is something that is documented as being done by Sendmail. There are two
7308 possible values. For messages generated by the autoreply transport, Exim
7311 Auto-Submitted: auto-replied
7313 whereas for all other generated messages (e.g. bounces) it adds
7315 Auto-Submitted: auto-generated
7317 63. The "personal" condition in filters now includes a test for the
7318 Auto-Submitted: header. If it contains the string "auto-" the message it
7319 not considered personal.
7321 64. Added rcpt_include_affixes as a generic transport option.
7323 65. Added queue_only_override (default true).
7325 66. Added the syslog_duplication option.
7327 67. If what should have been the first header line of a message consisted of
7328 a space followed by a colon, Exim was mis-interpreting it as a header line.
7329 It isn't of course - it is syntactically invalid and should therefore be
7330 treated as the start of the message body. The misbehaviour could have
7331 caused a number of strange effects, including loss of data in subsequent
7332 header lines, and spool format errors.
7334 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7335 client host had authenticated. This control can now be exercised by an ACL
7336 for more flexibility.
7338 69. By default, callouts do not happen when testing with -bh. There is now a
7339 variant, -bhc, which does actually run the callout code, including
7340 consulting and updating the callout cache.
7342 70. Added support for saslauthd authentication, courtesy of Alexander
7345 71. If statvfs() failed on the spool or log directories while checking their
7346 size for availability, Exim confusingly gave the error "space shortage".
7347 Furthermore, in debugging mode it crashed with a floating point exception.
7348 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7349 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7350 serious problem, Exim now writes to the main and panic logs when this
7351 happens, with details of the failure. It then refuses to accept the
7352 incoming message, giving the message "spool directory problem" or "log
7353 directory problem" with a 421 code for SMTP messages.
7355 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7356 0, 1, and 2 exist, because some OS complain for execs without them (see
7357 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7358 descriptors. However, the code omitted to check that the open succeeded,
7359 causing mysterious errors if for some reason the permissions on /dev/null
7360 got screwed. Now Exim writes a message to the main and panic logs, and
7361 bombs out if it can't open /dev/null.
7363 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7364 interact so that it is all more flexible. It is supposed to remain
7365 backwards compatible. Also added extra_local_interfaces.
7367 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7368 to bomb out with an assertion failure - to the client this appears as a
7369 connection drop. This problem occurs in the part of the code that was taken
7370 from the Samba project. Fortunately, the assertion is in a very simple
7371 function, so I have fixed this by reproducing the function inline in the
7372 one place where it is called, and arranging for authentication to fail
7373 instead of killing the process with assert().
7375 75. The SPA client code was not working when the server requested OEM rather
7376 than Unicode encoding.
7378 76. Added code to make require_files with a specific uid setting more usable in
7379 the case where statting the file as root fails - usually a non-root-mounted
7380 NFS file system. When this happens and the failure is EACCES, Exim now
7381 forks a subprocess and does the per-uid checking as the relevant uid.
7383 77. Added process_log_path.
7385 78. If log_file_path was not explicitly set, a setting of check_log_space or
7386 check_log_inodes was ignored.
7388 79. If a space check for the spool or log partitions fails, the incident is now
7389 logged. Of course, in the latter case the data may get lost...
7391 80. Added the %p formatting code to string_format() so that it can be used to
7392 print addresses in debug_print(). Adjusted all the address printing in the
7393 debugging in store.c to use %p rather than %d.
7395 81. There was a concern that a line of code in smtp_in.c could overflow a
7396 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7397 initially expressed, the concern was not well-founded, because trailing
7398 spaces are removed early. However, if the trailing spaces were followed by
7399 a NULL, they did not get removed, so the overflow was possible. Two fixes
7402 (a) I re-wrote the offending code in a cleaner fashion.
7403 (b) If an incoming SMTP command contains a NULL character, it is rejected
7406 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7407 runs initgroups(), so that if the Exim user is in any additional groups,
7408 they will be used during message reception.
7414 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob