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".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
77 JH/01 Remove code calling the customisable local_scan function, unless a new
78 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
80 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
81 non-signal-safe functions being used.
83 JH/03 Bug 2269: When presented with a received message having a stupidly large
84 number of DKIM-Signature headers, disable DKIM verification to avoid
85 a resource-consumption attack. The limit is set at twenty.
87 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
88 report of oldest_pass in ${authres } in consequence, and separate out
89 some descriptions of reasons for verification fail.
91 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
92 files in the spool were present and unlocked. A queue-runner could spot
93 them, resulting in a duplicate delivery. Fix that by doing the unlock
94 after the unlink. Investigation by Tim Stewart. Take the opportunity to
95 add more error-checking on spoolfile handling while that code is being
98 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
99 No known attacks, no CVE, this is defensive hardening.
101 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
102 a queue-runner could start a delivery while other operations were ongoing.
103 Cutthrough delivery was a common victim, resulting in duplicate delivery.
104 Found and investigated by Tim Stewart. Fix by using the open message data
105 file handle rather than opening another, and not locally closing it (which
106 releases a lock) for that case, while creating the temporary .eml format
107 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
109 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
110 $sender_verify_failure/$recipient_verify_failure to "random".
112 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
115 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
116 Previously this would segfault.
118 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
121 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
122 like zero, since the resolver should be doing this for us, But we need one
123 as a CNAME but no MX presence gets the CNAME returned; we need to check
124 that doesn't point to an MX to declare it "no MX returned" rather than
125 "error, loop". A new main option is added so the older capability of
126 following some limited number of chain links is maintained.
128 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
130 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
131 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
132 now-deprecated earlier definitions used only specified the range up to TLS
133 1.2 (in the older-version library docs).
135 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
137 JH/15 Rework TLS client-side context management. Stop using a global, and
138 explicitly pass a context around. This enables future use of TLS for
139 connections to service-daemons (eg. malware scanning) while a client smtp
140 connection is using TLS; with cutthrough connections this is quite likely.
142 JH/16 Fix ARC verification to do AS checks in reverse order.
144 JH/17 Support a "tls" option on the ${readsocket } expansion item.
146 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
147 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
148 Previously the "utf8" would be re-prepended for every additional message.
150 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
151 Previously thery were accepted, resulting in issues when attempting to
152 forward messages to a non-supporting MTA.
154 PP/02 Let -n work with printing macros too, not just options.
156 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
157 one parent address was copied, and bogus data was used at delivery-logging
158 time. Either a crash (after delivery) or bogus log data could result.
159 Discovery and analysis by Tim Stewart.
161 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
162 Previously if the string ended mid-character, we did not insert the
163 promised '?' replacement.
165 PP/04 Documentation: current string operators work on bytes, not codepoints.
167 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
168 should pack well giving a smaller memory footprint so better caching and
169 therefore performance. Group the declarations where this can't be done so
170 that the byte-sized flag variables are not interspersed among pointer
171 variables, giving a better chance of good packing by the compiler.
173 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
174 non-null, to avoid issues with sites running BATV. Previously reports were
175 sent with an empty envelope sender so looked like bounces.
177 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
178 The ignore_error flag wasn't being returned from the filter subprocess so
179 was not set for later routers. Investigation and fix by Matthias Kurz.
181 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
182 and a msg:complete for the whole, when a message is manually removed using
183 -Mrm. Developement by Matthias Kurz, hacked on by JH.
185 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
186 a "Gnu special" function, asprintf() in the DB utility binary builds; I
187 hope that is portable enough.
189 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
190 requiring a known-CA anchor certificate; make it now rely entirely on the
191 TLSA as an anchor. Checking the name on the leaf cert against the name
192 on the A-record for the host is still done for TA (but not for EE mode).
194 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
195 included in delivery lines for non-proxied connections, when compiled with
196 SUPPORT_SOCKS and running with proxy logging enabled.
198 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
199 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
200 move the existing event to fire before the normal logging of message
201 failure so that custom logging is bracketed by normal logging.
203 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
204 msg:fail:internal event. Developement by Matthias Kurz.
206 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
207 far too small for todays use of crypto signatures stored there. Go all
208 the way to the max DNS message size of 64kB, even though this might be
209 overmuch for IOT constrained device use.
211 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
212 copy a string over itself. The library routine is documented as not
213 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
215 JH/32 For main options check_spool_space and check_inode_space, where the
216 platform supports 64b integers, support more than the previous 2^31 kB
217 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
218 the previous G, M, k.
220 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
221 $authenticated_fail_id variable on authentication failure. Previously
224 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
225 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
226 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
227 GNUTLS_SEC_PARAM_MEDIUM.
229 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
230 side. Previously we would continue as if no SNI had been received.
232 JH/36 Harden the handling of string-lists. When a list consisted of a sole
233 "<" character, which should be a list-separator specification, we walked
234 off past the nul-terimation.
236 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
237 causes) even when the retry time is not yet met. Previously they were
238 not, meaning that when (say) an account was over-quota and temp-rejecting,
239 and multiple senders' messages were queued, only one sender would get
240 notified on each configured delay_warning cycle.
242 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
244 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
245 cipher-suites, an error can be left on the stack even for a succeeding
246 accept; this results in impossible error messages when a later operation
249 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
250 return error codes indicating retry. Under TLS1.3 this becomes required.
252 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
253 it only wrote the new authenticators, resulting in a lack of tracking of
254 peer changes of ESMTP extensions until the next cache flush.
256 JH/41 Fix the loop reading a message header line to check for integer overflow,
257 and more-often against header_maxsize. Previously a crafted message could
258 induce a crash of the recive process; now the message is cleanly rejected.
260 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
261 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
267 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
268 When redis_servers is set to a list of > 1 element, and the Redis servers
269 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
270 case of MOVED into a DEFER case instead, thus moving the query onto the
271 next server in the list. For a cluster of N elements, all N servers must
272 be defined in redis_servers.
274 GF/02 Catch and remove uninitialized value warning in exiqsumm
275 Check for existence of @ARGV before looking at $ARGV[0]
277 JH/01 Replace the store_release() internal interface with store_newblock(),
278 which internalises the check required to safely use the old one, plus
279 the allocate and data copy operations duplicated in both (!) of the
280 extant use locations.
282 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
283 modifier. This matches the restriction on the commandline.
285 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
286 Previously only the last row was returned.
288 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
289 we assumed that tags in the header were well-formed, and parsed the
290 element content after inspecting only the first char of the tag.
291 Assumptions at that stage could crash the receive process on malformed
294 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
295 While running the DKIM ACL we operate on the Permanent memory pool so that
296 variables created with "set" persist to the DATA ACL. Also (at any time)
297 DNS lookups that fail create cache records using the Permanent pool. But
298 expansions release any allocations made on the current pool - so a dnsdb
299 lookup expansion done in the DKIM ACL releases the memory used for the
300 DNS negative-cache, and bad things result. Solution is to switch to the
301 Main pool for expansions.
302 While we're in that code, add checks on the DNS cache during store_reset,
303 active in the testsuite.
304 Problem spotted, and debugging aided, by Wolfgang Breyha.
306 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
307 When none of the hosts presented to a transport match an already-open
308 connection, close it and proceed with the list. Previously we would
309 queue the message. Spotted by Lena with Yahoo, probably involving
312 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
313 Previously a spurious "250 OK id=" response was appended to the proper
316 JH/08 The "support for" informational output now, which built with Content
317 Scanning support, has a line for the malware scanner interfaces compiled
318 in. Interface can be individually included or not at build time.
320 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
321 by the template makefile "src/EDITME". The "STREAM" support for an older
322 ClamAV interface method is removed.
324 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
325 rows affected is given instead).
327 JH/11 The runtime Berkeley DB library version is now additionally output by
328 "exim -d -bV". Previously only the compile-time version was shown.
330 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
331 SMTP connection. Previously, when one had more receipients than the
332 first, an abortive onward connection was made. Move to full support for
333 multiple onward connections in sequence, handling cutthrough connection
334 for all multi-message initiating connections.
336 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
337 routers. Previously, a multi-recipient message would fail to match the
338 onward-connection opened for the first recipient, and cause its closure.
340 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
341 a timeout on read on a GnuTLS initiating connection, resulting in the
342 initiating connection being dropped. This mattered most when the callout
343 was marked defer_ok. Fix to keep the two timeout-detection methods
346 JH/15 Relax results from ACL control request to enable cutthrough, in
347 unsupported situations, from error to silently (except under debug)
348 ignoring. This covers use with PRDR, frozen messages, queue-only and
351 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
353 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
354 metadata, resulting in a crash in free().
356 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
357 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
358 Broken also in d185889f4, with init system revamp.
360 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
361 Previously we abruptly closed the connection after reading a malware-
362 found indication; now we go on to read the "scan ok" response line,
365 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
366 ACL. Previously, a crash would result.
368 JH/19 Speed up macro lookups during configuration file read, by skipping non-
369 macro text after a replacement (previously it was only once per line) and
370 by skipping builtin macros when searching for an uppercase lead character.
372 JH/20 DANE support moved from Experimental to mainline. The Makefile control
373 for the build is renamed.
375 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
376 was allocated for every new TLS startup, meaning one per message. Fix
377 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
379 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
380 reported the original. Fix to report (as far as possible) the ACL
381 result replacing the original.
383 JH/23 Fix memory leak during multi-message connections using STARTTLS under
384 OpenSSL. Certificate information is loaded for every new TLS startup,
385 and the resources needed to be freed.
387 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
389 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
392 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
393 DATA response info to the (existing) per-recipient response info for
394 the "C=" log element. It can have useful tracking info from the
395 destination system. Patch from Simon Arlott.
397 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
398 length value. Previously this would segfault.
400 HS/02 Support Avast multiline protoocol, this allows passing flags to
401 newer versions of the scanner.
403 JH/28 Ensure that variables possibly set during message acceptance are marked
404 dead before release of memory in the daemon loop. This stops complaints
405 about them when the debug_store option is enabled. Discovered specifically
406 for sender_rate_period, but applies to a whole set of variables.
407 Do the same for the queue-runner and queue-list loops, for variables set
408 from spool message files. Do the same for the SMTP per-message loop, for
409 certain variables indirectly set in ACL operations.
411 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
412 as a multi-recipient message from a mailinglist manager). The coding had
413 an arbitrary cutoff number of characters while checking for more input;
414 enforced by writing a NUL into the buffer. This corrupted long / fast
415 input. The problem was exposed more widely when more pipelineing of SMTP
416 responses was introduced, and one Exim system was feeding another.
417 The symptom is log complaints of SMTP syntax error (NUL chars) on the
418 receiving system, and refused recipients seen by the sending system
419 (propating to people being dropped from mailing lists).
420 Discovered and pinpointed by David Carter.
422 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
423 replaced by the ${authresults } expansion.
425 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
427 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
428 allows proper process termination in container environments.
430 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
431 Previously the "final dot" had a newline after it; ensure it is CR,LF.
433 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
434 and "err_perm", deprecated since 4.83 when the RFC-defined words
435 "temperror" and "permerror" were introduced.
437 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
438 transport-filters or DKIM-signing. The restriction was lost in the
439 consolidation of verify-callout and delivery SMTP handling.
440 Extend the restriction to also cover ARC-signing.
442 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
443 in defer=pass mode supply a 450 to the initiator. Previously the message
446 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
447 tls_require_ciphers is used as before.
449 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
450 "pass_unscanned". Only tmpfails from the scanner are written to
451 the paniclog, as they may require admin intervention (permission
452 denied, license issues). Other scanner errors (like decompression
453 bombs) do not cause a paniclog entry.
455 JH/36 Fix reinitialisation of DKIM logging variable between messages.
456 Previously it was possible to log spurious information in receive log
459 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
460 triggered odd behaviour from Outlook Express clients.
462 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
465 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
466 since the IETF WG has not yet settled on that versus the original
467 "bare" representation.
469 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
470 Previously the millisecond value corrupted the output.
471 Fix also for syslog_pid=no and log_selector +pid, for which the pid
472 corrupted the output.
478 JH/01 Rework error string handling in TLS interface so that the caller in
479 more cases is responsible for logging. This permits library-sourced
480 string to be attached to addresses during delivery, and collapses
481 pairs of long lines into single ones.
483 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
484 during configuration. Wildcards are allowed and expanded.
486 JH/02 Rework error string handling in DKIM to pass more info back to callers.
487 This permits better logging.
489 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
490 do not close it down and have the child transport start it up again on
491 the passed-on TCP connection. Instead, proxy the child (and any
492 subsequent ones) for TLS via a unix-domain socket channel. Logging is
493 affected: the continued delivery log lines do not have any DNSSEC, TLS
494 Certificate or OCSP information. TLS cipher information is still logged.
496 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
497 identical IP addresses on different listening ports. Will also affect
500 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
501 add noisy ifdef guards to special-case this sillyness.
502 Patch from Bernd Kuhls.
504 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
505 than 255 are no longer allowed.
507 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
508 Disable the session-cache too, which might reduce our load. Since we
509 currrectly use a new context for every connection, both as server and
510 client, there is no benefit for these.
511 GnuTLS appears to not support tickets server-side by default (we don't
512 call gnutls_session_ticket_enable_server()) but client side is enabled
513 by default on recent versions (3.1.3 +) unless the PFS priority string
516 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
517 <https://reproducible-builds.org/specs/source-date-epoch/>.
519 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
520 the check for any unsuccessful recipients did not notice the limit, and
521 erroneously found still-pending ones.
523 JH/08 Pipeline CHUNKING command and data together, on kernels that support
524 MSG_MORE. Only in-clear (not on TLS connections).
526 JH/09 Avoid using a temporary file during transport using dkim. Unless a
527 transport-filter is involved we can buffer the headers in memory for
528 creating the signature, and read the spool data file once for the
529 signature and again for transmission.
531 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
532 4.77 as the kernel support then wasn't solid, having issues in 64bit
533 mode. Now, it's been long enough. Add support for FreeBSD also.
535 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
536 case where the routing stage had gathered several addresses to send to
537 a host before calling the transport for the first, we previously failed
538 to close down TLS in the old transport process before passing the TCP
539 connection to the new process. The new one sent a STARTTLS command
540 which naturally failed, giving a failed delivery and bloating the retry
541 database. Investigation and fix prototype from Wolfgang Breyha.
543 JH/12 Fix check on SMTP command input synchronisation. Previously there were
544 false-negatives in the check that the sender had not preempted a response
545 or prompt from Exim (running as a server), due to that code's lack of
546 awareness of the SMTP input buffering.
548 PP/04 Add commandline_checks_require_admin option.
549 Exim drops privileges sanely, various checks such as -be aren't a
550 security problem, as long as you trust local users with access to their
551 own account. When invoked by services which pass untrusted data to
552 Exim, this might be an issue. Set this option in main configuration
553 AND make fixes to the calling application, such as using `--` to stop
556 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
557 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
558 responses to those, into a single TLS record each way (this usually means
559 a single packet). As a side issue, smtp_enforce_sync now works on TLS
562 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
563 affects you only if you're dancing at the edge of the param size limits.
564 If you are, and this message makes sense to you, then: raise the
565 configured limit or use OpenSSL 1.1. Nothing we can do for older
568 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
569 cmdline element to get the documented default one. Previously it was
572 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
573 are unsuable for non-message-specific reasons - eg. connection timeout,
574 banner-time rejection.
576 JH/16 Fix logging of delivery remote port, when specified by router, under
579 PP/06 Repair manualroute's ability to take options in any order, even if one
580 is the name of a transport.
583 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
585 JH/17 Change the list-building routines interface to use the expanding-string
586 triplet model, for better allocation and copying behaviour.
588 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
589 Previously it was constructed the first time a possibly-matching string
590 was met in the configuration file input during startup; now it is done
593 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
594 compatible one, to avoid the (poorly documented) possibility of a config
595 file in the working directory redirecting the DB files, possibly correpting
596 some existing file. CVE-2017-10140 assigned for BDB.
598 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
599 cache-hot. Previously, although the result was properly cached, the
600 initial verify call returned a defer.
602 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
603 the main verify for receipient in uncached-mode.
605 JH/22 Retire historical build files to an "unsupported" subdir. These are
606 defined as "ones for which we have no current evidence of testing".
608 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
609 if present. Previously it was ignored.
611 JH/24 Start using specified-initialisers in C structure init coding. This is
612 a C99 feature (it's 2017, so now considered safe).
614 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
615 if was a fixed-sized field and bitmask ops via macros; it is now more
618 PP/07 GitHub PR 56: Apply MariaDB build fix.
619 Patch provided by Jaroslav Škarvada.
621 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
622 during Coverity cleanups [4.87 JH/47]
623 Diagnosis and fix provided by Michael Fischer v. Mollard.
625 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
626 the right size to place the terminating semicolon on its own folded
627 line, the header hash was calculated to an incorrect value thanks to
628 the (relaxed) space the fold became.
630 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
631 and confused the parent.
633 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
634 which could crash as a result. This could lead to undeliverable messages.
636 JH/28 Logging: "next input sent too soon" now shows where input was truncated
639 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
640 matters on fast-turnover and PID-randomising systems, which were getting
641 out-of-order delivery.
643 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
644 a possibly-overlapping copy. The symptom was that "Remote host closed
645 connection in response to HELO" was logged instead of the actual 4xx
648 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
649 Previously only that bufferd was discarded, resulting in SYMTP command
652 JH/32 DKIM: when a message has multiple signatures matching an identity given
653 in dkim_verify_signers, run the dkim acl once for each. Previously only
654 one run was done. Bug 2189.
656 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
657 "panic the current process" to "deliberately defer". The panic log is
658 still written with the problem list name; the mail and reject logs now
659 get a temp-reject line for the message that was being handled, saying
660 something like "domains check lookup or other defer". The SMTP 451
661 message is still "Temporary local problem".
663 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
664 A crafted sequence of BDAT commands could result in in-use memory beeing
665 freed. CVE-2017-16943.
667 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
668 from SMTP input. Previously it was always done; now only done for DATA
669 and not BDAT commands. CVE-2017-16944.
671 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
672 to the message (such as an overlong header line). Previously this was
673 not done and we did not exit BDAT mode. Followon from the previous item
674 though a different problem.
680 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
681 than -2003 did; needs libidn2 in addition to libidn.
683 JH/02 The path option on a pipe transport is now expanded before use.
685 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
686 Patch provided by "Björn", documentation fix added too.
688 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
689 missing a wire-to-host endian conversion.
691 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
692 close after a BDAT command line could be taken as a following command,
693 giving a synch failure. Fix by only checking for synch immediately
694 before acknowledging the chunk.
696 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
697 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
698 macro. Patches provided by Josh Soref.
700 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
701 Previously we did not; the RFC seems ambiguous and VRFY is not listed
702 by IANA as a service extension. However, John Klensin suggests that we
705 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
706 the dkim code may be unix-mode line endings rather than smtp wire-format
707 CRLF, so prepend a CR to any bare LF.
709 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
710 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
712 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
713 the first were themselves being wrongly included in the feed into dkim
714 processing; with most chunk sizes in use this resulted in an incorrect
715 body hash calculated value.
717 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
718 DKIM signature block, for verification. Although advised against by
719 standards it is specifically not ruled illegal.
721 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
723 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
724 missing a body hash (the bh= tag).
726 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
727 It seems that HAProxy sends the Proxy Protocol information in clear and
728 only then does a TLS startup, so do the same.
730 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
731 TCP connections (such as for Spamd) unless the daemon successfully set
732 Fast Open mode on its listening sockets. This fixes breakage seen on
733 too-old kernels or those not configured for Fast Open, at the cost of
734 requiring both directions being enabled for TFO, and TFO never being used
735 by non-daemon-related Exim processes.
737 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
738 endings, at least on the first header line. Try to canonify any that get
739 past that check, despite the cost.
741 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
742 now limited to an arbitrary five deep, while parsing addresses with the
743 strip_excess_angle_brackets option enabled.
745 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
746 instead leave the unprompted TLS handshake in socket buffer for the
747 TLS library to consume.
749 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
751 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
753 JH/16 Drop variables when they go out of scope. Memory management drops a whole
754 region in one operation, for speed, and this leaves assigned pointers
755 dangling. Add checks run only under the testsuite which checks all
756 variables at a store-reset and panics on a dangling pointer; add code
757 explicitly nulling out all the variables discovered. Fixes one known
758 bug: a transport crash, where a dangling pointer for $sending_ip_address
759 originally assigned in a verify callout, is re-used.
761 PP/06 Drop '.' from @INC in various Perl scripts.
763 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
765 PP/08 Reduce a number of compilation warnings under clang; building with
766 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
767 should be warning-free.
769 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
771 HS/01 Fix portability problems introduced by PP/08 for platforms where
772 realloc(NULL) is not equivalent to malloc() [SunOS et al].
774 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
775 chunk. This allows us to accept broken chunked messages. We need a more
776 general solution here.
778 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
779 already-broken messages in the queue.
781 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
783 JH/19 Fix reference counting bug in routing-generated-address tracking.
789 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
790 supports it and a size is available (ie. the sending peer gave us one).
792 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
793 years of being deprecated). The replacements are the ACLs
794 acl_smtp_mime and acl_not_smtp_mime.
796 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
797 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
798 or even an in-clear connection were permitted. Now, if the host lookup
799 was dnssec and dane was requested then the host is only used if the
800 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
801 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
802 if one fails this test.
803 This means that a poorly-configured remote DNS will make it incommunicado;
804 but it protects against a DNS-interception attack on it.
806 JH/04 Bug 1810: make continued-use of an open smtp transport connection
807 non-noisy when a race steals the message being considered.
809 JH/05 If main configuration option tls_certificate is unset, generate a
810 self-signed certificate for inbound TLS connections.
812 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
813 in rewrites and routers.
815 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
816 and logged a warning sing 4.83; now they are a configuration file error.
818 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
819 (lacking @domain). Apply the same qualification processing as RCPT.
821 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
823 JH/10 Support ${sha256:} applied to a string (as well as the previous
826 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
827 a cutthrough deliver is pending, as we always want to make a connection.
828 This also avoids re-routing the message when later placing the cutthrough
829 connection after a verify cache hit.
830 Do not update it with the verify result either.
832 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
833 when routing results in more than one destination address.
835 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
836 signing (which inhibits the cutthrough capability). Previously only
837 the presence of an option was tested; now an expansion evaluating as
838 empty is permissible (obviously it should depend only on data available
839 when the cutthrough connection is made).
841 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
842 the relevant preceding SMTP command did not note the pipelining mode.
844 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
845 Previously they were not counted.
847 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
848 as one having no matching records. Previously we deferred the message
849 that needed the lookup.
851 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
852 distinguished as "(=".
854 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
855 for missing MX records. Previously it only worked for missing A records.
857 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
859 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
860 after the data-go-ahead and data-ack. Patch from Jason Betts.
862 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
863 even for a "none" policy. Patch from Tony Meyer.
865 JH/22 Fix continued use of a connection for further deliveries. If a port was
866 specified by a router, it must also match for the delivery to be
869 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
870 When one of the recipients of a message was unsuitable for the connection
871 (has no matching addresses), we lost track of needing to mark it
872 deferred. As a result mail would be lost.
874 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
876 JH/25 Decoding ACL controls is now done using a binary search; the source code
877 takes up less space and should be simpler to maintain. Merge the ACL
878 condition decode tables also, with similar effect.
880 JH/26 Fix problem with one_time used on a redirect router which returned the
881 parent address unchanged. A retry would see the parent address marked as
882 delivered, so not attempt the (identical) child. As a result mail would
885 JH/27 Fix a possible security hole, wherein a process operating with the Exim
886 UID can gain a root shell. Credit to http://www.halfdog.net/ for
887 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
890 JH/28 Enable {spool,log} filesystem space and inode checks as default.
891 Main config options check_{log,spool}_{inodes,space} are now
892 100 inodes, 10MB unless set otherwise in the configuration.
894 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
895 Previously it only applied to the main-section connection policy
898 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
900 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
901 by me. Added RFC7919 DH primes as an alternative.
903 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
904 are not in the system include path.
906 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
907 GnuTLS, when a session startup failed (eg because the client disconnected)
908 Exim did stdio operations after fclose. This was exposed by a recent
909 change which nulled out the file handle after the fclose.
911 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
912 signed directly by the cert-signing cert, rather than an intermediate
913 OCSP-signing cert. This is the model used by LetsEncrypt.
915 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
917 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
918 an incoming connection.
920 HS/02 Bug 1802: Do not half-close the connection after sending a request
923 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
924 fallback to "prime256v1".
926 JH/34 SECURITY: Use proper copy of DATA command in error message.
927 Could leak key material. Remotely exploitable. CVE-2016-9963.
933 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
934 and 3.4.4 - once the server is enabled to respond to an OCSP request
935 it does even when not requested, resulting in a stapling non-aware
936 client dropping the TLS connection.
938 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
939 support variable-length bit vectors. No functional change.
941 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
942 The I= interface field on outgoing lines is now after the H= remote
943 host field, same as incoming lines. There is a separate
944 outgoing_interface log selector which allows you to disable the
947 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
948 If not running log_selector +smtp_connection the mainlog would be held
949 open indefinitely after a "too many connections" event, including to a
950 deleted file after a log rotate. Leave the per net connection logging
951 leaving it open for efficiency as that will be quickly detected by the
952 check on the next write.
954 HS/01 Bug 1671: Fix post transport crash.
955 Processing the wait-<transport> messages could crash the delivery
956 process if the message IDs didn't exist for some reason. When
957 using 'split_spool_directory=yes' the construction of the spool
958 file name failed already, exposing the same netto behaviour.
960 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
961 mime_regex ACL conditions.
963 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
964 to DSN fail messages (bounces): remote IP, remote greeting, remote response
965 to HELO, local diagnostic string.
967 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
968 log line to debug. Even when configured with a tls authenticator many
969 client connections are expected to not authenticate in this way, so
970 an authenticate fail is not an error.
972 HS/02 Add the Exim version string to the process info. This way exiwhat
973 gives some more detail about the running daemon.
975 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
976 matter for fast-change records such as DNSBLs.
978 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
979 retry record, even if constant. There may be multiple transports with
980 different interface settings and the retry behaviour needs to be kept
983 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
985 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
987 JH/10 Bug 840: fix log_defer_output option of pipe transport
989 JH/11 Bug 830: use same host for all RCPTS of a message, even under
990 hosts_randomize. This matters a lot when combined with mua_wrapper.
992 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
993 ${quote_pgsql:<string>} operator.
995 JH/13 Bug 1708: avoid misaligned access in cached lookup.
997 JH/14 Change header file name for freeradius-client. Relevant if compiling
998 with Radius support; from the Gentoo tree and checked under Fedora.
1000 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1002 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1003 option transport_filter, meaning no filtering.
1005 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1007 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1008 defaults to "*" (all hosts). The variable is now available when not built
1009 with TLS, default unset, mainly to enable keeping the testsuite sane.
1010 If a server certificate is not supplied (via tls_certificate) an error is
1011 logged, and clients will find TLS connections fail on startup. Presumably
1012 they will retry in-clear.
1013 Packagers of Exim are strongly encouraged to create a server certificate
1014 at installation time.
1016 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1017 with the $config_file variable.
1019 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1020 in transport context, after the attempt, and per-recipient. The latter type
1021 is per host attempted. The event data is the error message, and the errno
1022 information encodes the lookup type (A vs. MX) used for the (first) host,
1023 and the trailing two digits of the smtp 4xx response.
1025 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1026 to write to mainlog (or rejectlog, paniclog) in the window between file
1027 creation and permissions/ownership being changed. Particularly affects
1028 installations where exicyclog is run as root, rather than exim user;
1029 result is that the running daemon panics and dies.
1031 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1033 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1034 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1035 "pri" and "weight". Note that the previous implicit priority given by the
1036 list order is no longer honoured.
1038 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1039 for DKIM processing.
1041 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1042 by defining SUPPORT_SOCKS.
1044 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1045 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1046 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1047 variables are renamed to proxy_{local,external}_{address,port}.
1049 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1050 for a build by defining SUPPORT_I18N
1052 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1053 of the query string, and make ${quote_redis:} do that quoting.
1055 JH/29 Move Events support from Experimental to mainline, enabled by default
1056 and removable for a build by defining DISABLE_EVENT.
1058 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1060 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1061 cached by the daemon.
1063 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1064 by defining LOOKUP_REDIS. The libhiredis library is required.
1066 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1067 keys are given for lookup.
1069 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1070 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1071 only supported when built with TLS support. The PolarSSL SHA routines
1072 are still used when the TLS library is too old for convenient support.
1074 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1075 openssl_options), for security. OpenSSL forces this from version 1.1.0
1076 server-side so match that on older versions.
1078 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1079 allocation for $value could be released as the expansion processing
1080 concluded, but leaving the global pointer active for it.
1082 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1083 and to use the domains and local_parts ACL conditions.
1085 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1086 incorrectly not doubled on cutthrough transmission, hence seen as a
1087 body-termination at the receiving system - resulting in truncated mails.
1088 Commonly the sender saw a TCP-level error, and retransmitted the message
1089 via the normal store-and-forward channel. This could result in duplicates
1090 received - but deduplicating mailstores were liable to retain only the
1091 initial truncated version.
1093 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1095 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1097 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1098 we're in there, support oversigning also; bug 1309.
1100 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1102 HS/04 Add support for keep_environment and add_environment options.
1104 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1105 either intentional arithmetic overflow during PRNG, or testing config-
1108 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1109 delivery resulted in actual delivery. Cancel cutthrough before DATA
1112 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1113 rejects a recipient: pass the reject to the originator.
1115 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1116 Many were false-positives and ignorable, but it's worth fixing the
1119 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1120 for the new environment-manipulation done at startup. Move the routines
1121 from being local to tls.c to being global via the os.c file.
1123 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1124 an extract embedded as result-arg for a map, the first arg for extract
1125 is unavailable so we cannot tell if this is a numbered or keyed
1126 extraction. Accept either.
1132 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1135 JH/02 The smtp transport option "multi_domain" is now expanded.
1137 JH/03 The smtp transport now requests PRDR by default, if the server offers
1140 JH/04 Certificate name checking on server certificates, when exim is a client,
1141 is now done by default. The transport option tls_verify_cert_hostnames
1142 can be used to disable this per-host. The build option
1143 EXPERIMENTAL_CERTNAMES is withdrawn.
1145 JH/05 The value of the tls_verify_certificates smtp transport and main options
1146 default to the word "system" to access the system default CA bundle.
1147 For GnuTLS, only version 3.0.20 or later.
1149 JH/06 Verification of the server certificate for a TLS connection is now tried
1150 (but not required) by default. The verification status is now logged by
1151 default, for both outbound TLS and client-certificate supplying inbound
1154 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1157 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1158 Status Notification (bounce) messages are now MIME format per RFC 3464.
1159 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1160 under the control of the dsn_advertise_hosts option, and routers may
1161 have a dsn_lasthop option.
1163 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1164 default, modifiable by a malware= option. The list separator for
1165 the options can now be changed in the usual way. Bug 68.
1167 JH/10 The smtp_receive_timeout main option is now expanded before use.
1169 JH/11 The incoming_interface log option now also enables logging of the
1170 local interface on delivery outgoing connections.
1172 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1173 if the interface and destination host and port all match.
1175 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1178 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1179 Patch from Andrew Lewis.
1181 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1182 now supports optional time-restrictions, weighting, and priority
1183 modifiers per server. Patch originally by <rommer@active.by>.
1185 JH/16 The spamd_address main option now supports a mixed list of local
1186 and remote servers. Remote servers can be IPv6 addresses, and
1187 specify a port-range.
1189 JH/17 Bug 68: The spamd_address main option now supports an optional
1190 timeout value per server.
1192 JH/18 Bug 1581: Router and transport options headers_add/remove can
1193 now have the list separator specified.
1195 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1198 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1201 JH/21 Support for the A6 type of dns record is withdrawn.
1203 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1204 rather than the verbs used.
1206 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1207 from 255 to 1024 chars.
1209 JH/24 Verification callouts now attempt to use TLS by default.
1211 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1212 are generic router options now. The defaults didn't change.
1214 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1215 Original patch from Alexander Shikoff, worked over by JH.
1217 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1218 files if ZCAT_COMMAND is not executable.
1220 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1222 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1224 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1225 Normally benign, it bites when the pair was led to by a CNAME;
1226 modern usage is to not canonicalize the domain to a CNAME target
1227 (and we were inconsistent anyway for A-only vs AAAA+A).
1229 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1231 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1232 when evaluating $sender_host_dnssec.
1234 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1235 $sender_helo_dnssec variable.
1237 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1239 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1241 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1243 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1244 documented as working, but never had. Support all but $spam_report.
1246 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1247 added for tls authenticator.
1249 HS/03 Add perl_taintmode main config option
1255 TL/01 When running the test suite, the README says that variables such as
1256 no_msglog_check are global and can be placed anywhere in a specific
1257 test's script, however it was observed that placement needed to be near
1258 the beginning for it to behave that way. Changed the runtest perl
1259 script to read through the entire script once to detect and set these
1260 variables, reset to the beginning of the script, and then run through
1261 the script parsing/test process like normal.
1263 TL/02 The BSD's have an arc4random API. One of the functions to induce
1264 adding randomness was arc4random_stir(), but it has been removed in
1265 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1266 function when detected.
1268 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1269 cause callback expansion.
1271 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1272 syntax errors in an expansion can be treated as a string instead of
1273 logging or causing an error, due to the internal use of bool_lax
1274 instead of bool when processing it.
1276 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1277 server certificates when making smtp deliveries.
1279 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1281 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1283 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1285 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1286 Merged patch from Sebastian Wiedenroth.
1288 JH/05 Fix results-pipe from transport process. Several recipients, combined
1289 with certificate use, exposed issues where response data items split
1290 over buffer boundaries were not parsed properly. This eventually
1291 resulted in duplicates being sent. This issue only became common enough
1292 to notice due to the introduction of connection certificate information,
1293 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1295 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1296 size buffer was used, resulting in syntax errors when an expansion
1299 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1300 version 3.3.6 or later.
1302 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1303 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1304 both become "event_action", the variables become $event_name, $event_data
1305 and $event_defer_errno. There is a new variable $verify_mode, usable in
1306 routers, transports and related events. The tls:cert event is now also
1307 raised for inbound connections, if the main configuration event_action
1310 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1311 early OCSP support, but no stapling (appears to be less than 1.0.0).
1313 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1314 server certificate names available under the smtp transport option
1315 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1318 JH/10 Time-related extraction expansions from certificates now use the main
1319 option "timezone" setting for output formatting, and are consistent
1320 between OpenSSL and GnuTLS compilations. Bug 1541.
1322 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1323 encoded parameter in the incoming message. Bug 1558.
1325 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1326 include certificate info, eximon was claiming there were spoolfile
1329 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1331 JH/14 Log delivery-related information more consistently, using the sequence
1332 "H=<name> [<ip>]" wherever possible.
1334 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1335 are problematic for Debian distribution, omit them from the release
1338 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1340 JH/16 Fix string representation of time values on 64bit time_t architectures.
1343 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1344 output list separator was used.
1349 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1350 checkers that were complaining about end of non-void function with no
1353 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1354 This was a regression introduced in 4.83 by another bugfix.
1356 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1358 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1359 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1365 TF/01 Correctly close the server side of TLS when forking for delivery.
1367 When a message was received over SMTP with TLS, Exim failed to clear up
1368 the incoming connection properly after forking off the child process to
1369 deliver the message. In some situations the subsequent outgoing
1370 delivery connection happened to have the same fd number as the incoming
1371 connection previously had. Exim would try to use TLS and fail, logging
1372 a "Bad file descriptor" error.
1374 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1375 utilities have not been installed.
1377 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1378 temporary space as the ACL may create new global variables.
1380 TL/01 LDAP support uses per connection or global context settings, depending
1381 upon the detected version of the libraries at build time.
1383 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1384 to extract and use the src ip:port in logging and expansions as if it
1385 were a direct connection from the outside internet. PPv2 support was
1386 updated based on HAProxy spec change in May 2014.
1388 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1390 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1391 Properly escape header and check for NULL return.
1393 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1396 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1398 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1399 characters in header names, implemented as a verify condition.
1400 Contributed by Michael Fischer v. Mollard.
1402 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1403 results permerror and temperror. Previous values are deprecated but
1404 still accepted. In a future release, err_perm and err_temp will be
1405 completely removed, which will be a backward incompatibility if the
1406 ACL tests for either of these two old results. Patch contributed by
1407 user bes-internal on the mailing list.
1409 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1411 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1412 selectors, in both main and reject logs.
1414 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1417 JH/07 Add malware type "sock" for talking to simple daemon.
1419 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1421 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1422 routers/transports under cutthrough routing.
1424 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1425 numbers. Touch up "bool" conditional to keep the same definition.
1427 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1429 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1431 JH/12 Expand items in router/transport headers_add or headers_remove lists
1432 individually rather than the list as a whole. Bug 1452.
1434 Required for reasonable handling of multiple headers_ options when
1435 they may be empty; requires that headers_remove items with embedded
1436 colons must have them doubled (or the list-separator changed).
1438 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1439 view the policy declared in the DMARC record. Currently, $dmarc_status
1440 is a combined value of both the record presence and the result of the
1443 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1445 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1446 dnslookup router and the smtp transport (applying to the forward
1449 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1450 of ldap servers used for a specific lookup. Patch provided by Heiko
1453 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1454 New variable $lookup_dnssec_authenticated for observability.
1456 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1457 Patch submitted by Lars Timman.
1459 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1461 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1462 Requires trusted mode and valid format message id, aborts otherwise.
1463 Patch contributed by Heiko Schlichting.
1465 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1466 certextract with support for various fields. Bug 1358.
1468 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1469 is requested by default, modifiable by smtp transport option
1472 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1473 operate on certificate variables to give certificate fingerprints
1474 Also new ${sha256:cert_variable}.
1476 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1478 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1479 Christian Aistleitner.
1481 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1483 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1484 file. Patch from Wolfgang Breyha.
1486 JH/25 Expand the coverage of the delivery $host and $host_address to
1487 client authenticators run in verify callout. Bug 1476.
1489 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1490 align with daemon_smtp_ports. Bug 72.
1492 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1493 support and error reporting did not work properly.
1495 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1496 and is readable. Patch from Andrew Colin Kissa.
1498 TL/14 Enhance documentation of ${run expansion and how it parses the
1499 commandline after expansion, particularly in the case when an
1500 unquoted variable expansion results in an empty value.
1502 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1504 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1505 by renaming away from it. Addresses GH issue 15, reported by
1508 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1510 TL/15 SECURITY: prevent double expansion in math comparison functions
1511 (can expand unsanitized data). Not remotely exploitable.
1518 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1520 PP/02 Make -n do something, by making it not do something.
1521 When combined with -bP, the name of an option is not output.
1523 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1526 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1527 $sender_host_name and config options to manage this, and basic check
1530 PP/05 DSCP support for outbound connections and control modifier for inbound.
1532 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1533 (Only plugin which currently uses this is kerberos4, which nobody should
1534 be using, but we should make it available and other future plugins might
1535 conceivably use it, even though it would break NAT; stuff *should* be
1536 using channel bindings instead).
1538 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1539 name; added for Sendmail compatibility; requires admin caller.
1540 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1541 just ignore it); requires trusted caller.
1542 Also parse but ignore: -Ac -Am -X<logfile>
1545 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1547 TL/02 Add +smtp_confirmation as a default logging option.
1549 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1550 Patch by Magnus Holmgren from 2007-02-20.
1552 TL/04 Bugzilla 1281 - Spec typo.
1553 Bugzilla 1283 - Spec typo.
1554 Bugzilla 1290 - Spec grammar fixes.
1556 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1558 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1560 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1561 mailing list by by Dmitry Isaikin.
1563 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1565 JH/02 Support "G" suffix to numbers in ${if comparisons.
1567 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1569 NM/01 Bugzilla 1197 - Spec typo
1570 Bugzilla 1196 - Spec examples corrections
1572 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1574 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1575 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1578 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1579 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1581 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1582 "acl {{name}{arg}...}", and optional args on acl condition
1585 JH/05 Permit multiple router/transport headers_add/remove lines.
1587 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1589 JH/07 Avoid using a waiting database for a single-message-only transport.
1590 Performance patch from Paul Fisher. Bugzilla 1262.
1592 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1595 JH/09 Add $headers_added variable, with content from use of ACL modifier
1596 add_header (but not yet added to the message). Bugzilla 199.
1598 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1599 Pulled from Bugzilla 817 by Wolfgang Breyha.
1601 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1603 (nb: this is the same fix as in Exim 4.80.1)
1605 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1608 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1611 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1613 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1614 advertise SMTP AUTH mechanism to us, instead of a generic
1615 protocol violation error. Also, make Exim more robust to bad
1616 data from the Dovecot auth socket.
1618 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1620 When a queue runner is handling a message, Exim first routes the
1621 recipient addresses, during which it prunes them based on the retry
1622 hints database. After that it attempts to deliver the message to
1623 any remaining recipients. It then updates the hints database using
1626 So if a recipient address works intermittently, it can get repeatedly
1627 deferred at routing time. The retry hints record remains fresh so the
1628 address never reaches the final cutoff time.
1630 This is a fairly common occurrence when a user is bumping up against
1631 their storage quota. Exim had some logic in its local delivery code
1632 to deal with this. However it did not apply to per-recipient defers
1633 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1635 This change adds a proper retry rule check during routing so that the
1636 final cutoff time is checked against the message's age. We only do
1637 this check if there is an address retry record and there is not a
1638 domain retry record; this implies that previous attempts to handle
1639 the address had the retry_use_local_parts option turned on. We use
1640 this as an approximation for the destination being like a local
1641 delivery, as in LMTP.
1643 I suspect this new check makes the old local delivery cutoff check
1644 redundant, but I have not verified this so I left the code in place.
1646 TF/02 Correct gecos expansion when From: is a prefix of the username.
1648 Test 0254 submits a message to Exim with the header
1652 When I ran the test suite under the user fanf2, Exim expanded
1653 the header to contain my full name, whereas it should have added
1654 a Resent-Sender: header. It erroneously treats any prefix of the
1655 username as equal to the username.
1657 This change corrects that bug.
1659 GF/01 DCC debug and logging tidyup
1660 Error conditions log to paniclog rather than rejectlog.
1661 Debug lines prefixed by "DCC: " to remove any ambiguity.
1663 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1665 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1666 Bug spotted by Jeremy Harris; was flawed since initial commit.
1667 Would have resulted in OCSP responses post-SNI triggering an Exim
1668 NULL dereference and crash.
1670 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1672 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1673 Bug detection, analysis and fix by Samuel Thibault.
1674 Bugzilla 1331, Debian bug #698092.
1676 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1678 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1679 Server implementation by Todd Lyons, client by JH.
1680 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1681 config variable "prdr_enable" controls whether the server
1682 advertises the facility. If the client requests PRDR a new
1683 acl_data_smtp_prdr ACL is called once for each recipient, after
1684 the body content is received and before the acl_smtp_data ACL.
1685 The client is controlled by both of: a hosts_try_prdr option
1686 on the smtp transport, and the server advertisement.
1687 Default client logging of deliveries and rejections involving
1688 PRDR are flagged with the string "PRDR".
1690 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1691 fclose(). Diagnosis by Todd Lyons.
1693 PP/17 Update configure.default to handle IPv6 localhost better.
1694 Patch by Alain Williams (plus minor tweaks).
1697 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1698 This is now consistent with GnuTLS, and is now documented: the
1699 previous undocumented portable approach to treating the option as
1700 unset was to force an expansion failure. That still works, and
1701 an empty string is now equivalent.
1703 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1704 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1705 not performing validation itself.
1707 PP/20 Added force_command boolean option to pipe transport.
1708 Patch from Nick Koston, of cPanel Inc.
1710 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1713 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1715 PP/21 Fix eximon continuous updating with timestamped log-files.
1716 Broken in a format-string cleanup in 4.80, missed when I repaired the
1717 other false fix of the same issue.
1718 Report and fix from Heiko Schlichting.
1721 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1722 Report from Prashanth Katuri.
1724 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1725 It's SecureTransport, so affects any MacOS clients which use the
1726 system-integrated TLS libraries, including email clients.
1728 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1729 using a MIME ACL for non-SMTP local injection.
1730 Report and assistance in diagnosis by Warren Baker.
1732 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1734 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1736 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1737 last id that failed so it may be referenced in subsequent ACL's.
1739 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1742 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1743 ldap library initialization, allowing self-signed CA's to be
1744 used. Also properly sets require_cert option later in code by
1745 using NULL (global ldap config) instead of ldap handle (per
1746 session). Bug diagnosis and testing by alxgomz.
1748 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1749 the src/util/ subdirectory.
1751 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1752 renamed to Transport Post Delivery Action by Jeremy Harris, as
1755 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1756 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1757 redis_servers = needs to be configured which will be used by the redis
1758 lookup. Patch from Warren Baker, of The Packet Hub.
1760 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1762 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1763 hostname or reverse DNS when processing a host list. Used suggestions
1764 from multiple comments on this bug.
1766 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1768 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1769 Missed a few lines, added it to make the runtest require no keyboard
1772 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1773 contains upper case chars. Make router use caseful_local_part.
1775 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1776 support when GnuTLS has been built with p11-kit.
1782 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1784 This, or similar/improved, will also be change PP/11 of 4.82.
1790 PP/01 Handle short writes when writing local log-files.
1791 In practice, only affects FreeBSD (8 onwards).
1792 Bugzilla 1053, with thanks to Dmitry Isaikin.
1794 NM/01 Bugzilla 949 - Documentation tweak
1796 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1799 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1801 PP/02 Implemented gsasl authenticator.
1803 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1805 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1806 `pkg-config foo` for cflags/libs.
1808 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1809 with rest of GSASL and with heimdal_gssapi.
1811 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1812 `pkg-config foo` for cflags/libs for the TLS implementation.
1814 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1815 properties get this fed in as external SSF. A number of robustness
1816 and debugging improvements to the cyrus_sasl authenticator.
1818 PP/08 cyrus_sasl server now expands the server_realm option.
1820 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1821 Patch by Jeremy Harris.
1823 PP/10 Added dbmjz lookup type.
1825 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1827 PP/12 MAIL args handles TAB as well as SP, for better interop with
1828 non-compliant senders.
1829 Analysis and variant patch by Todd Lyons.
1831 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1832 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1833 Patch from Dirk Mueller <dmueller@suse.com>
1835 PP/13 tls_peerdn now print-escaped for spool files.
1836 Observed some $tls_peerdn in wild which contained \n, which resulted
1837 in spool file corruption.
1839 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1840 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1841 or write after TLS renegotiation, which otherwise led to messages
1844 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1845 as a tracking header (ie: a signed header comes before the signature).
1846 Patch from Wolfgang Breyha.
1848 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1849 comma-sep list; embedded commas doubled.
1851 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1853 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1855 Report and patch from Dmitry Banschikov.
1857 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1858 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1859 failures. We appear to now support TLS1.1+ with Exim.
1861 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1862 lets Exim select keys and certificates based upon TLS SNI from client.
1863 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1864 before an outbound SMTP session. New log_selector, +tls_sni.
1866 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1867 NULL dereference. Report and patch from Alun Jones.
1869 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1870 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1871 Not seeing resolver debug output on NetBSD, but suspect this is a
1872 resolver implementation change.
1874 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1875 Left warnings. Added "eximon gdb" invocation mode.
1877 PP/21 Defaulting "accept_8bitmime" to true, not false.
1879 PP/22 Added -bw for inetd wait mode support.
1881 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1882 locate the relevant includes and libraries. Made this the default.
1884 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1885 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1887 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1888 This may cause build issues on older platforms.
1890 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1891 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1892 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1893 Added SNI support via GnuTLS too.
1894 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1896 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1898 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1899 Applied second patch from Janne, implementing suggestion to default
1900 multiple-strings-in-record handling to match SPF spec.
1902 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1904 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1905 read-only, out of scope).
1906 Patch from Wolfgang Breyha, report from Stuart Northfield.
1908 PP/29 Fix three issues highlighted by clang analyser static analysis.
1909 Only crash-plausible issue would require the Cambridge-specific
1910 iplookup router and a misconfiguration.
1911 Report from Marcin Mirosław.
1913 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1915 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1916 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1917 As part of this, removing so much warning spew let me fix some minor
1918 real issues in debug logging.
1920 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1921 assignment on my part. Fixed.
1923 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1924 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1925 Janne Snabb (who went above and beyond: thank you).
1927 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1928 string otherwise requires a connection and a bunch more work and it's
1929 relatively easy to get wrong. Should also expose TLS library linkage
1932 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1933 64-bit ${eval} (JH/03).
1935 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1936 GNU libc to support some of the 64-bit stuff, should not lead to
1937 conflicts. Defined before os.h is pulled in, so if a given platform
1938 needs to override this, it can.
1940 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1941 protection layer was required, which is not implemented.
1942 Bugzilla 1254, patch from Wolfgang Breyha.
1944 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1945 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1946 tls_dhparam take prime identifiers. Also unbreak combination of
1947 OpenSSL+DH_params+TLSSNI.
1949 PP/39 Disable SSLv2 by default in OpenSSL support.
1955 PP/01 Solaris build fix for Oracle's LDAP libraries.
1956 Bugzilla 1109, patch from Stephen Usher.
1958 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1960 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1963 TF/02 Fix a couple more cases where we did not log the error message
1964 when unlink() failed. See also change 4.74-TF/03.
1966 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1967 lock up or crash if it happened to be inside a call to libc when it
1968 got a SIGUSR1 from exiwhat.
1970 The SIGUSR1 handler appends the current process status to the process
1971 log which is later printed by exiwhat. It used to use the general
1972 purpose logging code to do this, but several functions it calls are
1973 not safe for signals.
1975 The new output code in the SIGUSR1 handler is specific to the process
1976 log, and simple enough that it's easy to inspect for signal safety.
1977 Removing some special cases also simplifies the general logging code.
1978 Removing the spurious timestamps from the process log simplifies
1981 TF/04 Improved ratelimit ACL condition.
1983 The /noupdate option has been deprecated in favour of /readonly which
1984 has clearer semantics. The /leaky, /strict, and /readonly update modes
1985 are mutually exclusive. The update mode is no longer included in the
1986 database key; it just determines when the database is updated. (This
1987 means that when you upgrade Exim will forget old rate measurements.)
1989 Exim now checks that the per_* options are used with an update mode that
1990 makes sense for the current ACL. For example, when Exim is processing a
1991 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1992 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1993 must specify per_mail/readonly. If you omit the update mode it defaults to
1994 /leaky where that makes sense (as before) or /readonly where required.
1996 The /noupdate option is now undocumented but still supported for
1997 backwards compatibility. It is equivalent to /readonly except that in
1998 ACLs where /readonly is required you may specify /leaky/noupdate or
1999 /strict/noupdate which are treated the same as /readonly.
2001 A useful new feature is the /count= option. This is a generalization
2002 of the per_byte option, so that you can measure the throughput of other
2003 aggregate values. For example, the per_byte option is now equivalent
2004 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2006 The per_rcpt option has been generalized using the /count= mechanism
2007 (though it's more complicated than the per_byte equivalence). When it is
2008 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2009 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2010 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2011 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2012 Note that using per_rcpt with a non-readonly update mode in more than
2013 one ACL will cause the recipients to be double-counted. (The per_mail
2014 and per_byte options don't have this problem.)
2016 The handling of very low rates has changed slightly. If the computed rate
2017 is less than the event's count (usually one) then this event is the first
2018 after a long gap. In this case the rate is set to the same as this event's
2019 count, so that the first message of a spam run is counted properly.
2021 The major new feature is a mechanism for counting the rate of unique
2022 events. The new per_addr option counts the number of different
2023 recipients that someone has sent messages to in the last time period. It
2024 behaves like per_rcpt if all the recipient addresses are different, but
2025 duplicate recipient addresses do not increase the measured rate. Like
2026 the /count= option this is a general mechanism, so the per_addr option
2027 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2028 example, measure the rate that a client uses different sender addresses
2029 with the options per_mail/unique=$sender_address. There are further
2030 details in the main documentation.
2032 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2034 TF/06 Removed a few PCRE remnants.
2036 TF/07 Automatically extract Exim's version number from tags in the git
2037 repository when doing development or release builds.
2039 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2040 Bugzilla 879. Patch from Paul Fisher.
2042 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2043 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2046 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2048 PP/05 Variable $av_failed, true if the AV scanner deferred.
2049 Bugzilla 1078. Patch from John Horne.
2051 PP/06 Stop make process more reliably on build failure.
2052 Bugzilla 1087. Patch from Heiko Schlittermann.
2054 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2055 Bugzilla 1089. Patch from Heiko Schlittermann.
2057 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2058 Bugzilla 1131. Patch from Holger Weiß.
2060 PP/09 Handle IPv6 addresses with SPF.
2061 Bugzilla 860. Patch from Wolfgang Breyha.
2063 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2065 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2068 PP/11 match_* no longer expand right-hand-side by default.
2069 New compile-time build option, EXPAND_LISTMATCH_RHS.
2070 New expansion conditions, "inlist", "inlisti".
2072 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2074 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2076 PP/14 fix log_write() format string regression from TF/03.
2077 Bugzilla 1152. Patch from Dmitry Isaikin.
2083 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2085 PP/02 Harmonised TLS library version reporting; only show if debugging.
2086 Layout now matches that introduced for other libraries in 4.74 PP/03.
2088 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2090 PP/04 New "dns_use_edns0" global option.
2092 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2095 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2096 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2098 TK/01 Updated PolarSSL code to 0.14.2.
2099 Bugzilla 1097. Patch from Andreas Metzler.
2101 PP/07 Catch divide-by-zero in ${eval:...}.
2102 Fixes bugzilla 1102.
2104 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2107 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2108 format-string attack -- SECURITY: remote arbitrary code execution.
2110 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2111 time unintentionally subject to list matching rules, letting the header
2112 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2113 arbitrary string expansion). This allowed for information disclosure.
2115 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2116 INT_MIN/-1 -- value coerced to INT_MAX.
2122 NM/01 Workaround for PCRE version dependency in version reporting
2125 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2126 This fixes portability to compilers other than gcc, notably
2127 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2129 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2130 makefiles for portability to HP-UX and POSIX correctness.
2132 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2133 Also via indented variable definition in the Makefile.
2134 (Debugging by Oliver Heesakkers).
2136 PP/02 Restore caching of spamd results with expanded spamd_address.
2137 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2139 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2140 Improves build reliability. Fix from: Frank Elsner
2142 NM/02 Fix wide character breakage in the rfc2047 coding
2143 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2145 NM/03 Allow underscore in dnslist lookups
2146 Fixes bug 1026. Patch from Graeme Fowler
2148 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2149 Code patches from Adam Ciarcinski of NetBSD.
2151 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2154 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2155 is logged, to avoid truncation. Patch from John Horne.
2157 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2158 Patch from Jakob Hirsch.
2160 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2161 SQL string expansion failure details.
2162 Patch from Andrey Oktyabrski.
2164 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2165 Patch from Simon Arlott.
2167 PP/09 New lookups functionality failed to compile on old gcc which rejects
2168 extern declarations in function scope.
2169 Patch from Oliver Fleischmann
2171 PP/10 Use sig_atomic_t for flags set from signal handlers.
2172 Check getgroups() return and improve debugging.
2173 Fixed developed for diagnosis in bug 927 (which turned out to be
2176 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2177 Patch from Mark Zealey.
2179 PP/12 Bugzilla 1056: Improved spamd server selection.
2180 Patch from Mark Zealey.
2182 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2183 Based on patch from Heiko Schlittermann.
2185 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2186 Patch from Uwe Doering, sign-off by Michael Haardt.
2188 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2189 variable declaration deep within a block. Bug and patch from
2192 PP/15 lookups-Makefile IRIX compatibility coercion.
2194 PP/16 Make DISABLE_DKIM build knob functional.
2196 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2197 Patch by Simon Arlott
2199 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2200 variable argument macros. Our copy now differs from upstream.
2206 TF/01 Failure to get a lock on a hints database can have serious
2207 consequences so log it to the panic log.
2209 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2210 controlled using the smtp_confirmation log selector.
2212 TF/03 Include the error message when we fail to unlink a spool file.
2214 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2215 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2216 for maintaining out-of-tree patches for some time.
2218 PP/01 Bugzilla 139: Documentation and portability issues.
2219 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2220 Handle per-OS dynamic-module compilation flags.
2222 PP/02 Let /dev/null have normal permissions.
2223 The 4.73 fixes were a little too stringent and complained about the
2224 permissions on /dev/null. Exempt it from some checks.
2225 Reported by Andreas M. Kirchwitz.
2227 PP/03 Report version information for many libraries, including
2228 Exim version information for dynamically loaded libraries. Created
2229 version.h, now support a version extension string for distributors
2230 who patch heavily. Dynamic module ABI change.
2232 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2233 privilege escalation vulnerability whereby the Exim run-time user
2234 can cause root to append content of the attacker's choosing to
2237 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2240 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2241 If dropping privileges for untrusted macros, we disabled normal logging
2242 on the basis that it would fail; for the Exim run-time user, this is not
2243 the case, and it resulted in successful deliveries going unlogged.
2244 Fixed. Reported by Andreas Metzler.
2250 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2251 only prepend for the Resent-* case. Fixes regression introduced in
2252 Exim 4.70 by NM/22 for Bugzilla 607.
2254 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2255 increasing numbers of administrators be bitten by this.
2257 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2259 PP/03 Bugzilla 994: added openssl_options main configuration option.
2261 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2263 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2265 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2267 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2268 without a peer certificate, leading to a segfault because of an
2269 assumption that peers always have certificates. Be a little more
2270 paranoid. Problem reported by Martin Tscholak.
2272 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2273 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2274 NB: ClamAV planning to remove STREAM in "middle of 2010".
2275 CL also introduces -bmalware, various -d+acl logging additions and
2276 more caution in buffer sizes.
2278 PP/09 Implemented reverse_ip expansion operator.
2280 PP/10 Bugzilla 937: provide a "debug" ACL control.
2282 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2284 PP/12 Bugzilla 973: Implement --version.
2286 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2288 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2290 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2292 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2293 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2294 ignore trailing whitespace.
2296 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2298 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2301 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2302 Notification from Dr Andrew Aitchison.
2304 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2305 ExtendedDetectionInfo response format.
2306 Notification from John Horne.
2308 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2311 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2312 XSL and documented dependency on system catalogs, with examples of how
2315 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2318 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2319 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2320 configuration file which is writeable by the Exim user or group.
2322 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2323 of configuration files to cover files specified with the -C option if
2324 they are going to be used with root privileges, not just the default
2327 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2328 option (effectively making it always true).
2330 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2331 files to be used while preserving root privileges.
2333 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2334 that rogue child processes cannot use them.
2336 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2337 run-time user, instead of root.
2339 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2340 Exim run-time user without dropping privileges.
2342 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2343 result string, instead of calling string_vformat() twice with the same
2346 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2347 for other users. Others should always drop root privileges if they use
2348 -C on the command line, even for a whitelisted configure file.
2350 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2352 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2358 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2359 $data_path, and $header_path variables; fixed documentation bugs and
2362 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2363 exipick to access non-standard spools, including the "frozen" queue
2366 NM/01 Bugzilla 965: Support mysql stored procedures.
2367 Patch from Alain Williams
2369 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2371 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2372 Patch from Andreas Metzler
2374 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2375 Patch from Kirill Miazine
2377 NM/05 Bugzilla 671: Added umask to procmail example.
2379 JJ/03 installed exipick 20100323.0, fixing doc bug
2381 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2382 directory. Notification and patch from Dan Rosenberg.
2384 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2386 TK/02 Improve log output when DKIM signing operation fails.
2388 MH/01 Treat the transport option dkim_domain as a colon separated
2389 list, not as a single string, and sign the message with each element,
2390 omitting multiple occurences of the same signer.
2392 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2393 Bugzilla 985, 986. Patch by Simon Arlott
2395 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2396 Patch by Simon Arlott
2398 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2399 MBX locking. Notification from Dan Rosenberg.
2405 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2407 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2409 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2411 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2413 NM/04 Bugzilla 915: Flag broken perl installation during build.
2419 TK/01 Added patch by Johannes Berg that expands the main option
2420 "spamd_address" if it starts with a dollar sign.
2422 TK/02 Write list of recipients to X-Envelope-Sender header when building
2423 the mbox-format spool file for content scanning (suggested by Jakob
2426 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2427 (http://www.dcc-servers.net/) support via dccifd. Activated by
2428 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2430 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2431 by Mark Daniel Reidel <mr@df.eu>.
2433 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2434 When building exim an external PCRE library is now needed -
2435 PCRE is a system library on the majority of modern systems.
2436 See entry on PCRE_LIBS in EDITME file.
2438 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2439 conversation. Added nologin parameter to request.
2440 Patch contributed by Kirill Miazine.
2442 TF/01 Do not log submission mode rewrites if they do not change the address.
2444 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2446 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2447 log files in place. Contributed by Roberto Lima.
2449 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2451 TF/03 Bugzilla 615: When checking the local_parts router precondition
2452 after a local_part_suffix or local_part_prefix option, Exim now
2453 does not use the address's named list lookup cache, since this
2454 contains cached lookups for the whole local part.
2456 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2457 Robert Millan. Documentation is in experimental-spec.txt.
2459 TF/04 Bugzilla 668: Fix parallel build (make -j).
2461 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2463 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2464 Patch provided by Jan Srzednicki.
2466 TF/05 Leading white space used to be stripped from $spam_report which
2467 wrecked the formatting. Now it is preserved.
2469 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2470 that they are available at delivery time.
2472 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2474 TF/08 TLS error reporting now respects the incoming_interface and
2475 incoming_port log selectors.
2477 TF/09 Produce a more useful error message if an SMTP transport's hosts
2478 setting expands to an empty string.
2480 NM/06 Bugzilla 744: EXPN did not work under TLS.
2481 Patch provided by Phil Pennock.
2483 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2484 Patch provided by Richard Godbee.
2486 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2487 acl_smtp_notquit, added index entry.
2489 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2490 Patch provided by Eugene Bujak.
2492 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2493 accept(). Patch provided by Maxim Dounin.
2495 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2496 Patch provided by Phil Pennock.
2498 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2500 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2501 Patch provided by Brad "anomie" Jorsch.
2503 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2504 Patch provided by Dean Brooks.
2506 TK/05 Add native DKIM support (does not depend on external libraries).
2508 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2509 Patch provided by Graeme Fowler.
2511 NM/16 Bugzilla 851: Documentation example syntax fix.
2513 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2515 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2518 NM/19 Bugzilla 745: TLS version reporting.
2519 Patch provided by Phil Pennock.
2521 NM/20 Bugzilla 167: bool: condition support.
2522 Patch provided by Phil Pennock.
2524 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2525 clients. Patch provided by Phil Pennock.
2527 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2528 Patch provided by Brad "anomie" Jorsch.
2530 NM/23 Bugzilla 687: Fix misparses in eximstats.
2531 Patch provided by Heiko Schlittermann.
2533 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2534 Patch provided by Heiko Schlittermann.
2536 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2537 plus update to original patch.
2539 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2541 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2542 Patch provided by David Brownlee.
2544 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2546 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2548 NM/30 Bugzilla 888: TLS documentation bugfixes.
2550 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2552 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2553 Unlike the original bugzilla I have changed all shell scripts in src tree.
2555 NM/33 Bugzilla 898: Transport filter timeout fix.
2556 Patch by Todd Rinaldo.
2558 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2559 Patch by Serge Demonchaux.
2561 NM/35 Bugzilla 39: Base64 decode bug fixes.
2562 Patch by Jakob Hirsch.
2564 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2566 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2568 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2570 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2576 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2577 ALT-N's libdkim that I have put here:
2578 http://duncanthrax.net/exim-experimental/
2580 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2581 were called 'true' and it seems that C99 defines that as a reserved
2582 keyword to be used with 'bool' variable types. That means you could
2583 not include C99-style headers which use bools without triggering
2584 build errors in sieve.c.
2586 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2587 as mailq or other aliases. Changed the --help handling significantly
2588 to do whats expected. exim_usage() emits usage/help information.
2590 SC/01 Added the -bylocaldomain option to eximstats.
2592 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2594 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2596 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2602 PH/01 Another patch from the Sieve maintainer.
2604 PH/02 When an IPv6 address is converted to a string for single-key lookup
2605 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2606 dots are used instead of colons so that keys in lsearch files need not
2607 contain colons. This was done some time before quoting was made available
2608 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2609 (notated using the quote facility) so as to distinguish them from IPv4
2610 keys. This meant that lookups for IP addresses in host lists did not work
2611 for iplsearch lookups.
2613 This has been fixed by arranging for IPv6 addresses to be expressed with
2614 colons if the lookup type is iplsearch. This is not incompatible, because
2615 previously such lookups could never work.
2617 The situation is now rather anomalous, since one *can* have colons in
2618 ordinary lsearch keys. However, making the change in all cases is
2619 incompatible and would probably break a number of configurations.
2621 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2624 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2625 conversion specification without a maximum field width, thereby enabling
2626 a rogue spamd server to cause a buffer overflow. While nobody in their
2627 right mind would setup Exim to query an untrusted spamd server, an
2628 attacker that gains access to a server running spamd could potentially
2629 exploit this vulnerability to run arbitrary code as the Exim user.
2631 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2632 $primary_hostname instead of what libspf2 thinks the hosts name is.
2634 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2635 a directory entry by the name of the lookup key. Previously, if a
2636 symlink pointed to a non-existing file or a file in a directory that
2637 Exim lacked permissions to read, a lookup for a key matching that
2638 symlink would fail. Now it is enough that a matching directory entry
2639 exists, symlink or not. (Bugzilla 503.)
2641 PH/03 The body_linecount and body_zerocount variables are now exported in the
2644 PH/04 Added the $dnslist_matched variable.
2646 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2647 This means they are set thereafter only if the connection becomes
2650 PH/06 Added the client_condition to authenticators so that some can be skipped
2651 by clients under certain conditions.
2653 PH/07 The error message for a badly-placed control=no_multiline_responses left
2654 "_responses" off the end of the name.
2656 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2658 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2659 (without spaces) instead of just copying the configuration text.
2661 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2663 PH/11 Added $max_received_linelength.
2665 PH/12 Added +ignore_defer and +include_defer to host lists.
2667 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2668 way in which PCRE > 7.0 is built.
2670 PH/14 Implemented queue_only_load_latch.
2672 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2673 MAIL command. The effect was to mangle the value on 64-bit systems.
2675 PH/16 Another patch from the Sieve maintainer.
2677 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2679 PH/18 If a system quota error occurred while trying to create the file for
2680 a maildir delivery, the message "Mailbox is full" was not appended to the
2681 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2682 only to a quota excession during the actual writing of the file.
2684 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2685 characters?) which causes problems in log lines. The DN values are now
2686 passed through string_printing() before being added to log lines.
2688 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2689 and InterBase are left for another time.)
2691 PH/21 Added message_body_newlines option.
2693 PH/22 Guard against possible overflow in moan_check_errorcopy().
2695 PH/23 POSIX allows open() to be a macro; guard against that.
2697 PH/24 If the recipient of an error message contained an @ in the local part
2698 (suitably quoted, of course), incorrect values were put in $domain and
2699 $local_part during the evaluation of errors_copy.
2705 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2706 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2709 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2710 issue a MAIL command.
2712 PH/02 In an ACL statement such as
2714 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2716 if a client was not listed at all, or was listed with a value other than
2717 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2718 the condition was not true (as it should be), so access was not denied.
2719 The bug was that the ! inversion was incorrectly passed on to the second
2720 item. This has been fixed.
2722 PH/03 Added additional dnslists conditions == and =& which are different from
2723 = and & when the dns lookup returns more than one IP address.
2725 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2726 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2728 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2729 FSYNC, which compiles an option called disable_fsync that allows for
2730 bypassing fsync(). The documentation is heavily laced with warnings.
2732 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2734 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2735 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2736 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2737 including adding "make clean"; (3) Added -fPIC when compiling the test
2738 dynamically loaded module, to get rid of a warning.
2740 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2741 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2742 = 0s. The bug is otherwise harmless.
2744 PH/07 There was a bug in the dovecot authenticator such that the value of
2745 $auth1 could be overwritten, and so not correctly preserved, after a
2746 successful authentication. This usually meant that the value preserved by
2747 the server_setid option was incorrect.
2749 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2751 PH/09 Installed PCRE release 7.0.
2753 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2754 run for batched SMTP input. It is now run at the start of every message
2755 in the batch. While fixing this I discovered that the process information
2756 (output by running exiwhat) was not always getting set for -bs and -bS
2757 input. This is fixed, and it now also says "batched" for BSMTP.
2759 PH/11 Added control=no_pipelining.
2761 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2762 patch, slightly modified), and move the expansion of helo_data till after
2763 the connection is made in the smtp transport (so it can use these
2766 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2768 PH/14 Added log_selector = +pid.
2770 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2772 PH/16 Add ${if forany and ${if forall.
2774 PH/17 Added dsn_from option to vary the From: line in DSNs.
2776 PH/18 Flush SMTP output before performing a callout, unless control =
2777 no_callout_flush is set.
2779 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2780 was true (the default) a successful delivery failed to delete the retry
2781 item, thus causing premature timeout of the address. The bug is now
2784 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2786 PH/21 Long custom messages for fakedefer and fakereject are now split up
2787 into multiline responses in the same way that messages for "deny" and
2788 other ACL rejections are.
2790 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2791 with slight modification.
2793 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2794 draft, changing the syntax and factoring some duplicate code".
2796 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2797 for deliveries of the second and subsequent messages over the same SMTP
2800 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2801 ${reduce, with only minor "tidies".
2803 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2805 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2806 expansion side effects.
2808 PH/27 When a message times out after an over-quota error from an Exim-imposed
2809 quota, the bounce message says "mailbox is full". This message was not
2810 being given when it was a system quota that was exceeded. It now should
2813 MH/03 Made $recipients available in local_scan(). local_scan() already has
2814 better access to the recipient list through recipients_list[], but
2815 $recipients can be useful in postmaster-provided expansion strings.
2817 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2818 in the case of a MAIL command with additional options following the
2819 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2820 were accidentally chopped off.
2822 PH/29 SMTP synchronization checks are implemented when a command is read -
2823 there is a check that no more input is waiting when there shouldn't be
2824 any. However, for some commands, a delay in an ACL can mean that it is
2825 some time before the response is written. In this time, more input might
2826 arrive, invalidly. So now there are extra checks after an ACL has run for
2827 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2828 pipelining has not been advertised.
2830 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2832 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2833 returned in the SMTP response when smtp_return_error_details was false.
2834 This has been fixed.
2836 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2837 instead of the C I/O that was originally supplied, because problems were
2838 reported on Solaris.
2840 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2841 Exim which did not show up earlier: it was assuming that a call to
2842 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2843 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2844 was a macro that became an assignment, so it seemed to work. This has
2845 changed to a proper function call with a void return, hence the compile
2846 error. Exim's code has been fixed.
2848 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2851 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2853 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2855 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2856 because it turns out to be too big or there is a timeout) there is no
2857 "Completed" line in the log. When some messages of this type were
2858 selected by exigrep, they were listed as "not completed". Others were
2859 picked up by some special patterns. I have improved the selection
2860 criteria to be more general.
2862 PH/38 The host_find_failed option in the manualroute router can now be set
2863 to "ignore", to completely ignore a host whose IP address cannot be
2864 found. If all hosts are ignored, the behaviour is controlled by the new
2865 host_all_ignored option.
2867 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2868 homing or because of multiple MX records with /mx) generated more than
2869 one IP address, and the following item turned out to be the local host,
2870 all the secondary addresses of the first item were incorrectly removed
2871 from the list, along with the local host and any following hosts (which
2872 is what is supposed to happen).
2874 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2875 whoever called Exim into the -H file. In the case of the daemon it was
2876 behaving confusingly. When first started, it used values for whoever
2877 started the daemon, but after a SIGHUP it used the Exim user (because it
2878 calls itself on a restart). I have changed the code so that it now always
2881 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2882 message are rejected with the same error (e.g. no authentication or bad
2883 sender address), and a DATA command is nevertheless sent (as can happen
2884 with PIPELINING or a stupid MUA), the error message that was given to the
2885 RCPT commands is included in the rejection of the DATA command. This is
2886 intended to be helpful for MUAs that show only the final error to their
2889 PH/42 Another patch from the Sieve maintainer.
2891 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2892 Eximstats - Fixed some broken HTML links and added missing column headers
2894 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2895 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2897 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2899 PH/43 Yet another patch from the Sieve maintainer.
2901 PH/44 I found a way to check for a TCP/IP connection going away before sending
2902 the response to the final '.' that terminates a message, but only in the
2903 case where the client has not sent further data following the '.'
2904 (unfortunately, this is allowed). However, in many cases there won't be
2905 any further data because there won't be any more messages to send. A call
2906 to select() can be used: if it shows that the input is "ready", there is
2907 either input waiting, or the socket has been closed. An attempt to read
2908 the next input character can distinguish the two cases. Previously, Exim
2909 would have sent an OK response which the client would never have see.
2910 This could lead to message repetition. This fix should cure that, at
2911 least in a lot of common cases.
2913 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2914 advertised in response to EHLO.
2920 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2921 fixed by 4.65/MH/01 (is this a record?) are fixed:
2923 (i) An empty string was always treated as zero by the numeric comparison
2924 operators. This behaviour has been restored.
2926 (ii) It is documented that the numeric comparison operators always treat
2927 their arguments as decimal numbers. This was broken in that numbers
2928 starting with 0 were being interpreted as octal.
2930 While fixing these problems I realized that there was another issue that
2931 hadn't been noticed. Values of message_size_limit (both the global option
2932 and the transport option) were treated as octal if they started with 0.
2933 The documentation was vague. These values are now always treated as
2934 decimal, and I will make that clear in the documentation.
2940 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2941 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2944 MH/01 Don't check that the operands of numeric comparison operators are
2945 integers when their expansion is in "skipping" mode (fixes bug
2946 introduced by 4.64-PH/07).
2948 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2949 child addresses, Exim now panics and dies. Previously, because the count
2950 is held in a short int, deliveries were likely to be lost. As such a
2951 large number of recipients for a single message is ridiculous
2952 (performance will be very, very poor), I have chosen to impose a limit
2953 rather than extend the field.
2959 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2960 leftover -K file (the existence of which was triggered by #402).
2961 While we were at it, introduced process PID as part of the -K
2962 filename. This should rule out race conditions when creating
2965 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2966 processing considerably. Previous code took too long for large mails,
2967 triggering a timeout which in turn triggers #401.
2969 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2970 in the DK code in transports.c. sendfile() is not really portable,
2971 hence the _LINUX specificness.
2973 TF/01 In the add_headers option to the mail command in an Exim filter,
2974 there was a bug that Exim would claim a syntax error in any
2975 header after the first one which had an odd number of characters
2978 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2979 callout verification, Exim cached a "reject" for the entire domain. This
2980 is correct for most verifications, but it is not correct for a recipient
2981 verification with use_sender or use_postmaster set, because in that case
2982 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2983 case of MAIL FROM:<> rejection from other early rejections (e.g.
2984 rejection of HELO). When verifying a recipient using a non-null MAIL
2985 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2986 Whatever the result of the callout, the value of the domain cache is
2987 left unchanged (for any other kind of callout, getting as far as trying
2988 RCPT means that the domain itself is ok).
2990 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2993 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2994 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2995 session, the daemon ignores ECONNECT errors and logs others; it now
2996 ignores EPIPE as well.
2998 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2999 (quoted-printable decoding).
3001 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3002 later the small subsequent patch to fix an introduced bug.
3004 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3006 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3008 PH/08 An error is now given if message_size_limit is specified negative.
3010 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3011 to be given (somewhat) arbitrary names.
3013 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3016 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3017 miscellaneous code fixes
3019 PH/10 Added the log_reject_target ACL modifier to specify where to log
3022 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3023 hostname. This is wrong, because it relates to the incoming message (and
3024 probably the interface on which it is arriving) and not to the outgoing
3025 callout (which could be using a different interface). This has been
3026 changed to use the value of the helo_data option from the smtp transport
3027 instead - this is what is used when a message is actually being sent. If
3028 there is no remote transport (possible with a router that sets up host
3029 addresses), $smtp_active_hostname is used.
3031 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3032 tweaks were necessary in order to get it to work (see also 21 below):
3033 (a) The code assumed that strncpy() returns a negative number on buffer
3034 overflow, which isn't the case. Replaced with Exim's string_format()
3036 (b) There were several signed/unsigned issues. I just did the minimum
3037 hacking in of casts. There is scope for a larger refactoring.
3038 (c) The code used strcasecmp() which is not a standard C function.
3039 Replaced with Exim's strcmpic() function.
3040 (d) The code set only $1; it now sets $auth1 as well.
3041 (e) A simple test gave the error "authentication client didn't specify
3042 service in request". It would seem that Dovecot has changed its
3043 interface. Fortunately there's a specification; I followed it and
3044 changed what the client sends and it appears to be working now.
3046 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3049 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3050 address A is aliased to B and C, where B exists and C does not. Without
3051 -v the output is "A verified" because verification stops after a
3052 successful redirection if more than one address is generated. However,
3053 with -v the child addresses are also verified. Exim was outputting "A
3054 failed to verify" and then showing the successful verification for C,
3055 with its parentage. It now outputs "B failed to verify", showing B's
3056 parentage before showing the successful verification of C.
3058 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3059 look up a TXT record in a specific list after matching in a combined
3062 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3063 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3064 they consult the DNS. I had assumed they would set it the way they
3065 wanted; and indeed my experiments on Linux seem to show that in some
3066 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3067 To be on the safe side, however, I have now made the interface to
3068 host_find_byname() similar to host_find_bydns(), with an argument
3069 containing the DNS resolver options. The host_find_byname() function now
3070 sets these options at its start, just as host_find_bydns() does. The smtp
3071 transport options dns_qualify_single and dns_search_parents are passed to
3072 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3073 of host_find_byname() use the default settings of RES_DEFNAMES
3074 (qualify_single) but not RES_DNSRCH (search_parents).
3076 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3077 spool_read_header() do less string testing, by means of a preliminary
3078 switch on the second character of optional "-foo" lines. (This is
3079 overdue, caused by the large number of possibilities that now exist.
3080 Originally there were few.) While I was there, I also converted the
3081 str(n)cmp tests so they don't re-test the leading "-" and the first
3082 character, in the hope this might squeeze out yet more improvement.
3084 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3085 flag allowing group syntax was set by the header_syntax check but not
3086 turned off, possible causing trouble later; (2) The flag was not being
3087 set at all for the header_verify test, causing "group"-style headers to
3088 be rejected. I have now set it in this case, and also caused header_
3089 verify to ignore an empty address taken from a group. While doing this, I
3090 came across some other cases where the code for allowing group syntax
3091 while scanning a header line wasn't quite right (mostly, not resetting
3092 the flag correctly in the right place). These bugs could have caused
3093 trouble for malformed header lines. I hope it is now all correct.
3095 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3096 with the "reply" argument non-NULL. The code, however (which originally
3097 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3098 but it didn't always do it. This confused somebody who was copying the
3099 code for some other use. I have removed all the tests.
3101 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3102 feature that was used to support insecure browsers during the U.S. crypto
3103 embargo. It requires special client support, and Exim is probably the
3104 only MTA that supported it -- and would never use it because real RSA is
3105 always available. This code has been removed, because it had the bad
3106 effect of slowing Exim down by computing (never used) parameters for the
3107 RSA_EXPORT functionality.
3109 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3110 authenticator to fail if there's a tab character in the incoming data
3111 (there should never be unless someone is messing about, as it's supposed
3112 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3113 if the connection is using TLS or if the remote IP is the same as the
3114 local IP, and the "valid-client-cert option" if a client certificate has
3117 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3118 authenticators. This can be used for authorization after authentication
3119 succeeds. (In the case of plaintext, it servers for both authentication
3122 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3123 if any retry times were supplied.
3125 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3126 connection, where there is no client IP address to check. In this
3127 situation, the verify now always succeeds.
3129 PH/25 Applied John Jetmore's -Mset patch.
3131 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3133 PH/27 In a string expansion for a processed (not raw) header when multiple
3134 headers of the same name were present, leading whitespace was being
3135 removed from all of them, but trailing whitespace was being removed only
3136 from the last one. Now trailing whitespace is removed from each header
3137 before concatenation. Completely empty headers in a concatenation (as
3138 before) are ignored.
3140 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3141 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3143 PH/29 [Removed. This was a change that I later backed out, and forgot to
3144 correct the ChangeLog entry (that I had efficiently created) before
3145 committing the later change.]
3147 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3148 address errors (4xx response to RCPT) over the same connection as other
3149 messages routed to the same hosts. Such deliveries are always "forced",
3150 so retry times are not inspected. This resulted in far too many retries
3151 for the affected addresses. The effect occurred only when there were more
3152 hosts than the hosts_max_try setting in the smtp transport when it had
3153 the 4xx errors. Those hosts that it had tried were not added to the list
3154 of hosts for which the message was waiting, so if all were tried, there
3155 was no problem. Two fixes have been applied:
3157 (i) If there are any address or message errors in an SMTP delivery, none
3158 of the hosts (tried or untried) are now added to the list of hosts
3159 for which the message is waiting, so the message should not be a
3160 candidate for sending over the same connection that was used for a
3161 successful delivery of some other message. This seems entirely
3162 reasonable: after all the message is NOT "waiting for some host".
3163 This is so "obvious" that I'm not sure why it wasn't done
3164 previously. Hope I haven't missed anything, but it can't do any
3165 harm, as the worst effect is to miss an optimization.
3167 (ii) If, despite (i), such a delivery is accidentally attempted, the
3168 routing retry time is respected, so at least it doesn't keep
3169 hammering the server.
3171 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3172 in ${readsocket because some servers need this prod.
3174 PH/32 Added some extra debug output when updating a wait-xxx database.
3176 PH/33 The hint "could be header name not terminated by colon", which has been
3177 given for certain expansion errors for a long time, was not being given
3178 for the ${if def:h_colon_omitted{... case.
3180 PH/34 The spec says: "With one important exception, whenever a domain list is
3181 being scanned, $domain contains the subject domain." There was at least
3182 one case where this was not true.
3184 PH/35 The error "getsockname() failed: connection reset by peer" was being
3185 written to the panic log as well as the main log, but it isn't really
3186 panic-worthy as it just means the connection died rather early on. I have
3187 removed the panic log writing for the ECONNRESET error when getsockname()
3190 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3191 runs only) independently of the message's sender address. This meant
3192 that, if the 4xx error was in fact related to the sender, a different
3193 message to the same recipient with a different sender could confuse
3194 things. In particular, this can happen when sending to a greylisting
3195 server, but other circumstances could also provoke similar problems.
3196 I have changed the default so that the retry time for these errors is now
3197 based a combination of the sender and recipient addresses. This change
3198 can be overridden by setting address_retry_include_sender=false in the
3201 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3202 remote server are returned as part of bounce messages. This was not
3203 happening for LMTP over a pipe (the lmtp transport), but now it is the
3204 same for both kinds of LMTP.
3206 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3207 in header lines that were in fact CNAMEs. This is no longer the case.
3209 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3210 and queue runs started by the daemon processed all messages. This has
3211 been fixed so that -R and -S can now usefully be given with -q<time>.
3213 PH/40 Import PCRE release 6.7 (fixes some bugs).
3215 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3217 PH/42 Give an error if -q is specified more than once.
3219 PH/43 Renamed the variables $interface_address and $interface_port as
3220 $received_ip_address and $received_port, to make it clear that these
3221 values apply to message reception, and not to the outgoing interface when
3222 a message is delivered. (The old names remain recognized, of course.)
3224 PH/44 There was no timeout on the connect() call when using a Unix domain
3225 socket in the ${readsocket expansion. There now is.
3227 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3228 be meaningful with "accept".
3230 SC/01 Eximstats V1.43
3231 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3233 SC/02 Eximstats V1.44
3234 Use a glob alias rather than an array ref in the generated
3235 parser. This improves both readability and performance.
3237 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3238 Collect SpamAssassin and rejection statistics.
3239 Don't display local sender or destination tables unless
3240 there is data to show.
3241 Added average volumes into the top table text output.
3243 SC/04 Eximstats V1.46
3244 Collect data on the number of addresses (recipients)
3245 as well as the number of messages.
3247 SC/05 Eximstats V1.47
3248 Added 'Message too big' to the list of mail rejection
3249 reasons (thanks to Marco Gaiarin).
3251 SC/06 Eximstats V1.48
3252 Mainlog lines which have GMT offsets and are too short to
3253 have a flag are now skipped.
3255 SC/07 Eximstats V1.49 (Alain Williams)
3256 Added the -emptyok flag.
3258 SC/08 Eximstats V1.50
3259 Fixes for obtaining the IP address from reject messages.
3261 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3262 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3263 whitespace changes from 4.64-PH/27
3265 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3268 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3269 are found, allow negative numbers in numeric criteria)
3271 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3273 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3274 to match changes made in 4.64-PH/43
3276 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3278 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3279 "domain" is actually an IP address, and force a failure. This locks out
3280 those revolvers/nameservers that support "A-for-A" lookups, in
3281 contravention of the specifications.
3283 PH/48 When a host name was looked up from an IP address, and the subsequent
3284 forward lookup of the name timed out, the host name was left in
3285 $sender_host_name, contrary to the specification.
3287 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3288 restricted to single-key lookups, Exim was not diagnosing an error if
3289 * or *@ was used with a query-style lookup.
3291 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3293 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3294 long ago, but noone interested enough thought of it. Let's just say that
3295 the "1.1" means that there are some new functions that weren't there at
3296 some point in the past.
3298 PH/51 Error processing for expansion failure of helo_data from an smtp
3299 transport during callout processing was broken.
3301 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3302 tested/used via the -bh/-bhc/-bs options.
3304 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3305 bug, fixed in subsequent PCRE releases).
3307 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3308 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3310 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3316 SC/01 Use a glob alias rather than an array ref in eximstats generated
3317 parser. This improves both readability and performance.
3319 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3320 Don't display local sender or destination tables in eximstats unless
3321 there is data to show.
3322 Added average volumes into the eximstats top table text output.
3324 SC/03 Collect data on the number of addresses (recipients) as well
3325 as the number of messages in eximstats.
3327 TF/01 Correct an error in the documentation for the redirect router. Exim
3328 does (usually) call initgroups() when daemonizing.
3330 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3331 with consistent privilege compared to when running as a daemon.
3333 TF/03 Note in the spec that $authenticated_id is not set for local
3334 submissions from trusted users.
3336 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3337 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3339 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3340 by adding some example configuration directives to the default
3341 configuration file. A little bit of work is required to uncomment the
3342 directives and define how usernames and passwords are checked, but
3343 there is now a framework to start from.
3345 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3346 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3347 without this. I don't know how relevant this is to other LDAP libraries.
3349 PH/02 Add the verb name to the "unknown ACL verb" error.
3351 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3353 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3355 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3356 directory not expanded when it should be if an expanded home directory
3357 was set for the address (which is overridden by the transport).
3359 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3362 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3363 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3364 because it is too late at that time, and has no effect.
3366 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3367 security issue with \' (bugzilla #107). I could not use the
3368 PQescapeStringConn() function, because it needs a PGconn value as one of
3371 PH/08 When testing addresses using -bt, indicate those final addresses that
3372 are duplicates that would not cause an additional delivery. At least one
3373 person was confused, thinking that -bt output corresponded to deliveries.
3374 (Suppressing duplicates isn't a good idea as you lose the information
3375 about possibly different redirections that led to the duplicates.)
3377 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3378 systems where poll() doesn't work, in particular OS X.
3380 PH/10 Added more information to debugging output for retry time not reached.
3382 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3383 operations in malware.c.
3385 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3388 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3389 syslog_duplication set false, log lines that would normally be written
3390 both the the main log and to the reject log were not written to syslog at
3393 PH/14 In the default configuration, change the use of "message" in ACL warn
3394 statements to "add_header".
3396 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3397 not followed by a command (e.g. "seen endif").
3399 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3400 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3403 PH/17 Added extra conditions to the default value of delay_warning_condition
3407 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3408 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3409 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3412 The Auto-Submitted: and various List- headers are standardised, whereas I
3413 don't think Precedence: ever was.
3415 PH/18 Refactored debugging code in route_finduser() to show more information,
3416 in particular, the error code if getpwnam() issues one.
3418 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3419 This is apparently needed in addition to the PH/07 change above to avoid
3420 any possible encoding problems.
3422 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3423 but not after initializing Perl.
3425 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3426 output them only if debugging. By default they are written stderr,
3427 apparently, which is not desirable.
3429 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3432 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3435 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3437 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3438 authenticated or an ident call has been made. Suppress the default
3439 values for $authenticated_id and $authenticated_sender (but permit -oMai
3440 and -oMas) when testing with -bh.
3442 PH/24 Re-jigged the order of the tests in the default configuration so that the
3443 tests for valid domains and recipients precede the DNS black list and CSA
3444 tests, on the grounds that those ones are more expensive.
3446 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3447 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3448 This bug exists in every version of Exim that I still have, right back to
3451 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3452 However, an attempt to turn on case-sensitivity in a regex key by
3453 including (?-i) didn't work because the subject string was already
3454 lowercased, and the effects were non-intuitive. It turns out that a
3455 one-line patch can be used to allow (?-i) to work as expected.
3461 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3462 other effects) broke the use of negated acl sub-conditions.
3464 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3467 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3468 "Deny" causes Exim to reject the incoming connection with a 554 error.
3469 Unfortunately, if there is a major crisis, such as a disk failure,
3470 tcp-wrappers gives "deny", whereas what one would like would be some
3471 kind of temporary error. A kludge has been added to help with this.
3472 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3473 554 error is used if errno is still zero or contains ENOENT (which occurs
3474 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3477 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3479 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3480 errors. Otherwise a message that provokes a temporary error (when other
3481 messages do not) can cause a whole host to time out.
3483 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3484 addresses were routed directly to files or pipes from a redirect router.
3485 File deliveries just didn't batch; pipe deliveries might have suffered
3488 PH/06 A failure to get a lock for a hints database would erroneously always say
3489 "Failed to get write lock", even when it was really a read lock.
3491 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3492 of 0600. This has been changed to use the value of the lockfile_mode
3493 option (which defaults to 0600).
3495 PH/08 Applied small patch from the Sieve maintainer.
3497 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3498 folder from quota calculations, a direct delivery into this folder messed
3499 up the contents of the maildirsize file. This was because the regex was
3500 used only to exclude .Trash (or whatever) when the size of the mailbox
3501 was calculated. There was no check that a delivery was happening into an
3502 excluded directory. This bug has been fixed by ignoring all quota
3503 processing for deliveries into excluded directories.
3505 PH/10 Added the maildirfolder_create_regex option to appendfile.
3511 PH/01 The code for finding all the local interface addresses on a FreeBSD
3512 system running IPv6 was broken. This may well have applied to all BSD
3513 systems, as well as to others that have similar system calls. The broken
3514 code found IPv4 interfaces correctly, but gave incorrect values for the
3515 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3516 that it would not match correctly against @[] and not recognize the IPv6
3519 PH/02 The ipliteral router was not recognizing addresses of the form user@
3520 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3522 PH/03 Added disable_ipv6.
3524 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3525 decoded form, because it is most often used to construct To: headers
3526 lines in autoreplies, and the decoded form may well be syntactically
3527 invalid. However, $reply_address has leading white space removed, and all
3528 newlines turned into spaces so that the autoreply transport does not
3531 PH/05 If group was specified without a user on a router, and no group or user
3532 was specified on a transport, the group from the router was ignored.
3534 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3535 for visible compile-time settings that can be used to change these
3536 numbers, for those that want even more. Backwards compatibility with old
3537 spool files has been maintained. However, going back to a previous Exim
3538 release will lost any variables that are in spool files.
3540 PH/07 Two small changes when running in the test harness: increase delay when
3541 passing a TCP/IP connection to a new process, in case the original
3542 process has to generate a bounce, and remove special handling of
3543 127.0.0.2 (sic), which is no longer necessary.
3545 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3546 be the same on different OS.
3548 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3551 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3552 whether --show-vars was specified or not
3554 JJ/02 exipick: Added support for new ACL variable spool format introduced
3557 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3558 syntactically invalid From: or Reply-to: line, and a filter used this to
3559 generate an autoreply, and therefore failed to obtain an address for the
3560 autoreply, Exim could try to deliver to a non-existent relative file
3561 name, causing unrelated and misleading errors. What now happens is that
3562 it logs this as a hard delivery error, but does not attempt to create a
3565 PH/11 The exinext utility has a -C option for testing purposes, but although
3566 the given file was scanned by exinext itself; it wasn't being passed on
3567 when Exim was called.
3569 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3570 an end-of-file indication when reading a command response.
3572 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3573 compiled. In many other places in Exim, IPv6 addresses are always
3574 recognized, so I have changed this. It also means that IPv4 domain
3575 literals of the form [IPV4:n.n.n.n] are now always recognized.
3577 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3578 used if the router is not running as root, for example, when verifying at
3579 ACL time, or when using -bh. The debugging output from this situation was
3580 non-existent - all you got was a failure to exec. I have made two
3583 (a) Failures to set uid/gid, the current directory, or a process leader
3584 in a subprocess such as that created by queryprogram now generate
3585 suitable debugging output when -d is set.
3587 (b) The queryprogram router detects when it is not running as root,
3588 outputs suitable debugging information if -d is set, and then runs
3589 the subprocess without attempting to change uid/gid.
3591 PH/15 Minor change to Makefile for building test_host (undocumented testing
3594 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3595 additional section of a DNS packet that returns MX or SRV records.
3596 Instead, it always explicitly searches for A/AAAA records. This avoids
3597 major problems that occur when a DNS server includes only records of one
3598 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3599 fixed another bug: if SRV records were looked up and the corresponding
3600 address records were *not* found in the additional section, the port
3601 values from the SRV records were lost.
3603 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3604 using the correct key (the original address) when searching the retry
3605 rules in order to find which one to use for generating the retry hint.
3607 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3608 adding the default one. Similarly, if it contains a Reply-To: header, the
3609 errors_reply_to option, if set, is not used.
3611 PH/19 When calculating a retry time, Exim used to measure the "time since
3612 failure" by looking at the "first failed" field in the retry record. Now
3613 it does not use this if it is later than than the arrival time of the
3614 message. Instead it uses the arrival time. This makes for better
3615 behaviour in cases where some deliveries succeed, thus re-setting the
3616 "first failed" field. An example is a quota failure for a huge message
3617 when small messages continue to be delivered. Without this change, the
3618 "time since failure" will always be short, possible causing more frequent
3619 delivery attempts for the huge message than are intended.
3620 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3622 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3623 $1, $2, $3) because the numerical variables can be reset during some
3624 expansion items (e.g. "match"), thereby losing the authentication data.
3626 PH/21 Make -bV show the size of off_t variables so that the test suite can
3627 decide whether to run tests for quotas > 2G.
3629 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3630 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3631 filecount value is greater than 2G or if a quota value is greater than 2G
3632 on a system where the size of off_t is not greater than 4, a panic error
3635 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3636 never match. The debug and -bh output now contains an explicit error
3637 message indicating a malformed IPv4 address or mask.
3639 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3640 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3641 PH/23 above applies.
3643 PH/25 Do not write to syslog when running in the test harness. The only
3644 occasion when this arises is a failure to open the main or panic logs
3645 (for which there is an explicit test).
3647 PH/26 Added the /no_tell option to "control=freeze".
3649 PH/27 If a host name lookup failed very early in a connection, for example, if
3650 the IP address matched host_lookup and the reverse lookup yielded a name
3651 that did not have a forward lookup, an error message of the form "no IP
3652 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3653 could be logged. Now it outputs the IP address instead of "NULL".
3655 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3656 allows the sender and the authenticated sender to be set when
3657 submitting a message from within Exim. Since child_open_exim() is
3658 documented for local_scan(), the new function should be too.
3660 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3661 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3662 results in an empty string is now treated as unset.
3664 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3666 PH/31 Added sender_verify_fail logging option.
3668 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3669 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3670 tidied the source and removed it altogether.
3672 PH/33 When a queue run was abandoned because the load average was too high, a
3673 log line was always written; now it is written only if the queue_run log
3674 selector is set. In addition, the log line for abandonment now contains
3675 information about the queue run such as the pid. This is always present
3676 in "start" and "stop" lines but was omitted from the "abandon" line.
3678 PH/34 Omit spaces between a header name and the colon in the error message that
3679 is given when verify = headers_syntax fails (if there are lots of them,
3680 the message gets confusing).
3682 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3683 names, as there are now some PTR records that contain slashes. This check
3684 is only to protect against broken name servers that fall over on strange
3685 characters, so the fact that it applies to all lookups doesn't matter.
3687 PH/36 Now that the new test suite is complete, we can remove some of the
3688 special code in Exim that was needed for the old test suite. For example,
3689 sorting DNS records because real resolvers return them in an arbitrary
3690 order. The new test suite's fake resolver always returns records in the
3693 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3694 bounces) except when queue_only is set, to avoid logging races between
3695 the different processes.
3697 PH/38 Panic-die if .include specifies a non-absolute path.
3699 PH/39 A tweak to the "H" retry rule from its user.
3701 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3702 a label. They prevented compilation on older perls.
3704 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3705 a warning to be raised on newish perls.
3707 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3708 on queue. Changes to match documented behaviour of showing count of
3709 messages matching specified criteria.
3711 PH/40 Changed the default ident timeout from 30s to 5s.
3713 PH/41 Added support for the use of login_cap features, on those BSD systems
3714 that have them, for controlling the resources used by pipe deliveries.
3716 PH/42 The content-scanning code uses fopen() to create files in which to put
3717 message data. Previously it was not paying any attention to the mode of
3718 the files. Exim runs with umask(0) because the rest of the code creates
3719 files with open(), and sets the required mode explicitly. Thus, these
3720 files were ending up world-writeable. This was not a big issue, because,
3721 being within the spool directory, they were not world-accessible. I have
3722 created a function called modefopen, which takes an additional mode
3723 argument. It sets umask(777), creates the file, chmods it to the required
3724 mode, then resets the umask. All the relevant calls to fopen() in the
3725 content scanning code have been changed to use this function.
3727 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3728 to 24 hours. This avoids potential overflow problems when processing G
3729 and H retry rules. I suspect nobody ever tinkers with this value.
3731 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3733 PH/45 When the plaintext authenticator is running as a client, the server's
3734 challenges are checked to ensure they are valid base64 strings. By
3735 default, the authentication attempt is cancelled if an invalid string is
3736 received. Setting client_ignore_invalid_base64 true ignores these errors.
3737 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3738 they are received. Thus, the responses can be made to depend on the
3739 challenges. If an invalid string is ignored, an empty string is placed in
3742 PH/46 Messages that are created by the autoreply transport now contains a
3743 References: header, in accordance with RFCs 2822 and 3834.
3745 PH/47 Added authenticated_sender_force to the smtp transport.
3747 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3749 PH/49 Installed latest patch from the Sieve maintainer.
3751 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3752 was also set, the appendfile transport was unnecessarily scanning a
3753 directory of message files (e.g. for maildir delivery) to find the count
3754 of files (along with the size), even though it did not need this
3755 information. It now does the scan only if it needs to find either the
3756 size of the count of files.
3758 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3760 PH/52 Two bugs concerned with error handling when the smtp transport is
3763 (i) Exim was not creating retry information for temporary errors given
3764 for individual recipients after the DATA command when the smtp transport
3765 was used in LMTP mode. This meant that they could be retried too
3766 frequently, and not timed out correctly.
3768 (ii) Exim was setting the flag that allows error details to be returned
3769 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3770 recipients that were returned after the DATA command.
3772 PH/53 This is related to PH/52, but is more general: for any failing address,
3773 when detailed error information was permitted to be returned to the
3774 sender, but the error was temporary, then after the final timeout, only
3775 "retry timeout exceeded" was returned. Now it returns the full error as
3776 well as "retry timeout exceeded".
3778 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3779 do this, and (what is worse) MTAs that accept it.
3781 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3782 will now be deprecated.
3784 PH/56 New os.c-cygwin from the Cygwin maintainer.
3786 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3787 formats (previously only available in exim formats via -bpr, -bpru,
3788 and -bpra. Now also available in native and exiqgrep formats)
3790 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3791 with very large, slow to parse queues
3793 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3795 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3797 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3798 responses to authentication challenges, though it was showing the
3799 challenges; (ii) I've removed the CR characters from the debug output for
3802 PH/58 Allow for the insertion of a newline as well as a space when a string
3803 is turned into more than one encoded-word during RFC 2047 encoding. The
3804 Sieve code now uses this.
3806 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3807 data_4xx, lost_connection, tls_required.
3809 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3810 message was being sent as an SMTP response.
3812 PH/61 Add -l and -k options to exicyclog.
3814 PH/62 When verifying, if an address was redirected to one new address, so that
3815 verification continued, and the new address failed or deferred after
3816 having set something in $address_data, the value of $address_data was not
3817 passed back to the ACL. This was different to the case when no
3818 redirection occurred. The value is now passed back in both cases.
3820 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3821 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3822 use login_cap.h, so on its own it isn't the distinguishing feature. The
3823 new name refers directly to the setclassresources() function.
3825 PH/65 Added configuration files for NetBSD3.
3827 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3829 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3830 is preferred over IPv4.
3832 PH/68 The bounce_return_message and bounce_return_body options were not being
3833 honoured for bounces generated during the reception of non-SMTP messages.
3834 In particular, this applied to messages rejected by the ACL. This bug has
3835 been fixed. However, if bounce_return_message is true and bounce_return_
3836 body is false, the headers that are returned for a non-SMTP message
3837 include only those that have been read before the error was detected.
3838 (In the case of an ACL rejection, they have all been read.)
3840 PH/69 The HTML version of the specification is now built in a directory called
3841 spec_html instead of spec.html, because the latter looks like a path with
3842 a MIME-type, and this confuses some software.
3844 PH/70 Catch two compiler warnings in sieve.c.
3846 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3847 function verify_get_ident() calls ip_connect() to connect a socket, but
3848 if the "connect()" function timed out, ip_connect() used to close the
3849 socket. However, verify_get_ident() also closes the socket later, and in
3850 between Exim writes to the log, which may get opened at this point. When
3851 the socket was closed in ip_connect(), the log could get the same file
3852 descriptor number as the socket. This naturally causes chaos. The fix is
3853 not to close the socket in ip_connect(); the socket should be closed by
3854 the function that creates it. There was only one place in the code where
3855 this was missing, in the iplookup router, which I don't think anybody now
3856 uses, but I've fixed it anyway.
3858 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3859 well as to direct DNS lookups. Otherwise the handling of names in host
3860 lists is inconsistent and therefore confusing.
3866 PH/01 Two changes to the default runtime configuration:
3868 (1) Move the checks for relay_from_hosts and authenticated clients from
3869 after to before the (commented out) DNS black list checks.
3871 (2) Add control=submission to the relay_from_hosts and authenticated
3872 clients checks, on the grounds that messages accepted by these
3873 statements are most likely to be submissions.
3875 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3877 (1) Generate an error if the third argument for the ${prvs expansion is
3880 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3883 (3) Reset the variables that are obtained from the first argument of
3884 ${prvscheck and used in the second argument before leaving the code,
3885 because their memory is reclaimed, so using them afterwards may do
3888 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3889 one (it's much easier than Tom thought :-).
3891 (5) Because of (4), we can now allow for the use of $prvscheck_result
3892 inside the third argument.
3894 PH/03 For some reason, the default setting of PATH when running a command from
3895 a pipe transport was just "/usr/bin". I have changed it to
3898 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3899 anything to be listed in the output from -bV.
3901 PH/05 When a filter generated an autoreply, the entire To: header line was
3902 quoted in the delivery log line, like this:
3904 => >A.N.Other <ano@some.domain> <original@ddress> ...
3906 This has been changed so that it extracts the operative address. There
3907 may be more than one such address. If so, they are comma-separated, like
3910 => >ano@some.domain,ona@other.domain <original@ddress> ...
3912 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3913 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3914 not being looked up in the rDNS to get a host name, Exim was showing the
3915 IP address twice in Received: lines, even though the IP addresses were
3916 identical. For example:
3918 Received: from [1.2.3.4] (helo=[1.2.3.4])
3920 However, if the real host name was known, it was omitting the HELO data
3921 if it matched the actual IP address. This has been tidied up so that it
3922 doesn't show the same IP address twice.
3924 PH/07 When both +timestamp and +memory debugging was on, the value given by
3925 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3926 called by the debug printing, thereby overwriting the timestamp buffer.
3927 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3929 PH/08 When the original message was included in an autoreply transport, it
3930 always said "this is a copy of the message, including all the headers",
3931 even if body_only or headers_only was set. It now gives an appropriate
3934 PH/09 Applied a patch from the Sieve maintainer which:
3936 o fixes some comments
3937 o adds the (disabled) notify extension core
3938 o adds some debug output for the result of if/elsif tests
3939 o points to the current vacation draft in the documentation
3940 and documents the missing references header update
3944 o fixes a bug in processing the envelope test (when testing
3945 multiple envelope elements, the last element determined the
3948 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3949 Electronic Mail") by including:
3951 Auto-submitted: auto-generated
3953 in the messages that it generates (bounce messages and others, such as
3954 warnings). In the case of bounce messages for non-SMTP messages, there was
3955 also a typo: it was using "Auto_submitted" (underscore instead of
3956 hyphen). Since every message generated by Exim is necessarily in response
3957 to another message, thes have all been changed to:
3959 Auto-Submitted: auto-replied
3961 in accordance with these statements in the RFC:
3963 The auto-replied keyword:
3965 - SHOULD be used on messages sent in direct response to another
3966 message by an automatic process,
3968 - MUST NOT be used on manually-generated messages,
3970 - MAY be used on Delivery Status Notifications (DSNs) and Message
3971 Disposition Notifications (MDNs),
3973 - MUST NOT be used on messages generated by automatic or periodic
3974 processes, except for messages which are automatic responses to
3977 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3978 to the default Received: header definition.
3980 PH/12 Added log selector acl_warn_skipped (default on).
3982 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3983 variables because (a) they are in the wrong storage pool and (b) even if
3984 they were copied, it wouldn't work properly because of the caching.
3986 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3987 checking when decoding. Apparently there are clients that generate
3988 overlong encoded strings. Why am I not surprised?
3990 PH/15 If the first argument of "${if match_address" was not empty, but did not
3991 contain an "@" character, Exim crashed. Now it writes a panic log message
3992 and treats the condition as false.
3994 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3996 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3997 extension "envelope-auth". The code is finished and in agreement with
3998 other implementations, but there is no documentation so far and in fact,
3999 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4000 not changing the active code.
4002 Print executed "if" and "elsif" statements when debugging is used. This
4003 helps a great deal to understand what a filter does.
4005 Document more things not specified clearly in RFC3028. I had all this
4006 sorted out, when out of a sudden new issues came to my mind. Oops."
4008 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4011 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4012 canonical form (as documented). However, after a host name lookup from
4013 the IP address, check_host() was doing a simple string comparison with
4014 addresses acquired from the DNS when checking that the found name did
4015 have the original IP as one of its addresses. Since any found IPv6
4016 addresses are likely to be in abbreviated form, the comparison could
4017 fail. Luckily, there already exists a function for doing the comparison
4018 by converting both addresses to binary, so now that is used instead of
4019 the text comparison.
4021 PH/20 There was another similar case to PH/19, when a complete host name was
4022 given in a host list; looking up its IP address could give an abbreviated
4023 form, whereas the current host's name might or might not be abbreviated.
4024 The same fix has been applied.
4030 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4031 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4034 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4035 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4037 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4039 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4040 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4041 Also, for "Auto-Submitted", treat anything other than "no" as
4042 non-personal, in accordance with RFC 3834. (Previously it treated
4043 anything starting "auto-" as non-personal.)
4045 TF/01 The control=submission/name=... option had a problem with syntax
4046 errors if the name included a slash character. The /name= option
4047 now slurps the rest of the string, so it can include any characters
4048 but it must come last in the list of options (after /sender_retain
4051 PH/05 Some modifications to the interface to the fake nameserver for the new
4059 TK/01 Added the "success_on_redirect" address verification option. See
4060 NewStuff for rationale and an example.
4062 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4064 PH/02 Patch to exigrep to allow it to work on syslog lines.
4066 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4067 fread() to skip over the body file's header line, because in Cygwin the
4068 header line is locked and is inaccessible.
4070 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4071 co-exist for some time) to make it clear that it is the Exim ID that is
4072 referenced, not the Message-ID: header line.
4074 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4075 string_format() function, because snprintf() does not exist on all
4078 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4079 ${stat: expansion item.
4081 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4082 protocol synchronization error", to keep the pedants happy.
4084 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4085 well as for IRIX systems, when gcc is being used. See the host.c source
4088 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4090 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4093 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4094 transport and to the smtp transport in LMTP mode.
4096 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4098 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4099 run a filter in a subprocess. This could lead to confusion in subsequent
4100 lookups in the parent process. There should also be a search_tidyup() at
4101 the end of the subprocess.
4103 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4104 only if the host matched helo_try_verify_hosts, which caused the
4105 verification to occur when the EHLO/HELO command was issued. The ACL just
4106 tested the remembered result. Now, if a previous verification attempt has
4107 not happened, "verify = helo" does it there and then.
4109 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4111 TK/03 Fix log output including CR from clamd.
4113 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4114 exist provoked a memory error which could cause a segfault.
4116 PH/15 Installed PCRE 6.2
4118 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4120 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4121 of the problem. Specifically, suggested +O2 rather than +O1 for the
4124 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4126 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4127 record, the port was not correctly logged when the outgoing_port log
4128 selector was set (it logged the transort's default port).
4130 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4131 fallback_hosts, and "hosts" in the smtp transport.
4133 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4134 host errors such as "Connection refused".
4136 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4137 authentication with radiusclient 0.4.9:
4139 - Error returned from rc_read_config was caught wrongly
4140 - Username/password not passed on to radius server due to wrong length.
4142 The presumption is that some radiusclient API changes for 4.51/PH/17
4143 were not taken care of correctly. The code is still untested by me (my
4144 Linux distribution still has 0.3.2 of radiusclient), but it was
4145 contributed by a Radius user.
4147 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4148 expanding the "port" option of the smtp transport.
4150 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4151 while reading a MIME header. Thanks to Tom Hughes for a patch.
4153 PH/24 Include config.h inside local_scan.h so that configuration settings are
4156 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4157 that in an ACL for RCPT (for example), you can examine exactly what was
4160 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4161 commands, but it was not correctly comparing the address with the actual
4162 client host address. Thus, it would show the EHLO address in Received:
4163 header lines when this was not necessary.
4165 PH/27 Added the % operator to ${eval:}.
4167 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4168 it should be ignoring failures (because with -C, for example, it has lost
4169 privilege). It wasn't ignoring creation failures other than "already
4172 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4175 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4176 because an input line was too long, either on its own, or by virtue of
4177 too many continuations, the temporary file was not being removed, and the
4178 return code was incorrect.
4180 PH/31 Missing "BOOL" in function definition in filtertest.c.
4182 PH/32 Applied Sieve patches from the maintainer.
4184 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4186 PH/33 Added "verify = not_blind".
4188 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4189 Local/Makefile (with some defaults set). These are used in built scripts
4190 such as exicyclog, but they have never been used in the exim_install
4191 script (though there are many overriding facilities there). I have
4192 arranged that the exim_install script now takes note of these two
4195 PH/35 Installed configuration files for Dragonfly.
4197 PH/36 When a locally submitted message by a trusted user did not contain a
4198 From: header, and the sender address was obtained from -f or from an SMTP
4199 MAIL command, and the trusted user did not use -F to supply a sender
4200 name, $originator_name was incorrectly used when constructing a From:
4201 header. Furthermore, $originator_name was used for submission mode
4202 messages from external hosts without From: headers in a similar way,
4203 which is clearly wrong.
4205 PH/37 Added control=suppress_local_fixups.
4207 PH/38 When log_selector = +received_sender was set, and the addition of the
4208 sender made the log line's construction buffer exactly full, or one byte
4209 less than full, an overflow happened when the terminating "\n" was
4212 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4213 when the result of a list match is failure because a DNS lookup failed.
4215 PH/40 RM_COMMAND is now used in the building process.
4217 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4218 the "build-* directories that it finds.
4220 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4221 address in a domain literal was a prefix of an interface address.
4223 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4224 when verifying a sender address, unless rewrite_headers is false.
4226 PH/44 Wrote a long comment about why errors_to addresses are verified as
4227 recipients, not senders.
4229 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4230 the ratelimit ACL was added.
4232 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4234 PH/46 Added extra information about PostgreSQL errors to the error string.
4236 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4237 avoiding the need to install special zones in a real server. This is
4238 backwards compatible; if it can't find the fake resolver, it drops back.
4239 Thus, both old and new test suites can be run.
4241 TF/02 Added util/ratelimit.pl
4243 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4244 clock is set back in time.
4246 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4247 Candler <B.Candler@pobox.com>.
4249 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4250 OK for addresses that are the result of redirections.
4252 PH/48 A number of further additions for the benefit of the new test suite,
4253 including a fake gethostbyname() that interfaces to the fake DNS resolver
4256 TF/06 The fix for widen_domains has also been applied to qualify_single and
4257 search_parents which are the other dnslookup options that can cause
4260 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4263 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4265 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4266 service" if it encountered an MX record with an empty target hostname.
4267 The message is now "an MX or SRV record indicated no SMTP service".
4269 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4270 if the DNS of the sending site is misconfigured. This is quite a
4271 common situation. This change restores the behaviour of treating a
4272 helo verification defer as a failure.
4274 PH/51 If self=fail was set on a router, the bounce message did not include the
4275 actual error message.
4281 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4283 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4284 command itself ended in error, the underlying message about the transport
4285 filter timeout was being overwritten with the pipe command error. Now the
4286 underlying error message should be appended to the second error message.
4288 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4290 PH/02 Reduce the amount of output that "make" produces by default. Full output
4291 can still be requested.
4293 PH/03 The warning log line about a condition test deferring for a "warn" verb
4294 was being output only once per connection, rather than after each
4295 occurrence (because it was using the same function as for successful
4296 "warn" verbs). This seems wrong, so I have changed it.
4298 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4299 it should not have, which might have caused a crash in the right
4300 circumstances, but probably never did.
4302 PH/04 Installed a modified version of Tony Finch's patch to make submission
4303 mode fix the return path as well as the Sender: header line, and to
4304 add a /name= option so that you can make the user's friendly name appear
4307 TF/03 Added the control = fakedefer ACL modifier.
4309 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4310 Mark Lowes for thorough testing.
4312 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4314 TK/03 Merged latest SRS patch from Miles Wilton.
4316 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4317 with the definition in sysexits.h (which is #included earlier).
4318 Fortunately, Exim does not actually use EX_OK. The code used to try to
4319 preserve the sysexits.h value, by assuming that macro definitions were
4320 scanned for macro replacements. I have been disabused of this notion,
4321 so now the code just undefines EX_OK before #including unistd.h.
4323 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4324 in the smtp transport. When a block could not be written in a single
4325 write() function, the timeout was being re-applied to each part-write.
4326 This seems wrong - if the receiver was accepting one byte at a time it
4327 would take for ever. The timeout is now adjusted when this happens. It
4328 doesn't have to be particularly precise.
4330 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4331 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4333 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4334 without a domain if the check to <postmaster@domain> fails.
4336 SC/01 Eximstats: added -xls and the ability to specify output files
4337 (patch written by Frank Heydlauf).
4339 SC/02 Eximstats: use FileHandles for outputting results.
4341 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4343 SC/04 Eximstats: fixed display of large numbers with -nvr option
4345 SC/05 Eximstats: fixed merging of reports with empty tables.
4347 SC/06 Eximstats: added the -include_original_destination flag
4349 SC/07 Eximstats: removed tabs and trailing whitespace.
4351 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4353 TK/06 MBOX spool code: Add real "From " MBOX separator line
4354 so the .eml file is really in mbox format (even though
4355 most programs do not really care). Patch from Alex Miller.
4357 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4358 The latter is generated from $received_to and is only set if the
4359 message has one envelope recipient. SA can use these headers,
4360 obviously out-of-the-box. Patch from Alex Miller.
4362 PH/08 The ${def test on a variable was returning false if the variable's
4363 value was "0", contrary to what the specification has always said!
4364 The result should be true unless the variable is empty.
4366 PH/09 The syntax error of a character other than { following "${if
4367 def:variable_name" (after optional whitespace) was not being diagnosed.
4368 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4369 accidental colon was present, for example, could give incorrect results.
4371 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4372 result is used (not including appendfile, where other changes are about
4375 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4376 This involved changing a lot of size variables from int to off_t. It
4377 should work with maildirs and everything.
4379 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4380 spamd dying while we are connected to it.
4382 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4385 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4386 function for BDB 4.3.
4388 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4390 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4391 the sources for PCRE have been re-arranged and more files are now
4394 PH/15 The code I had for printing potentially long long variables in PH/11
4395 above was not the best (it lost precision). The length of off_t variables
4396 is now inspected at build time, and an appropriate printing format (%ld
4397 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4398 to be "long long int" or "long int". This is needed for the internal
4399 formatting function string_vformat().
4401 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4402 the configuration file to be ":syslog", then the script "guesses" where
4403 the logs files are, rather than using the compiled in default. In our
4404 case the guess is not the same as the compiled default, so the script
4405 suddenly stopped working when I started to use syslog. The patch checks
4406 to see if log_file_path is "". If so, it attempts to read it from exim
4407 with no configuration file to get the compiled in version, before it
4408 falls back to the previous guessing code."
4410 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4411 implementing BATV in an Exim configuration. See NewStuff for the gory
4414 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4415 Makefile that are specific to HP-UX.
4417 PH/18 If the "use_postmaster" option was set for a recipient callout together
4418 with the "random" option, the postmaster address was used as the MAIL
4419 FROM address for the random test, but not for the subsequent recipient
4420 test. It is now used for both.
4422 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4423 patch removes a few documentation additions to RFC 3028, because the
4424 latest draft now contains them. It adds the new en;ascii-case comparator
4425 and a new error check for 8bit text in MIME parts. Comparator and
4426 require names are now matched exactly. I enabled the subaddress
4427 extension, but it is not well tested yet (read: it works for me)."
4429 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4430 rework some of the code of TK/09 above to avoid the hardwired use of
4431 "%lld" and "long long". Replaced the call to snprintf() with a call to
4434 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4435 records point to non-existent hosts", "retry timeout exceeded", and
4436 "retry time not reached for any host after a long failure period".
4438 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4439 experimental DomainKeys support:
4441 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4442 (2) On an error such as an illegally used "control", the wrong name for
4443 the control was given.
4445 These problems did NOT occur unless DomainKeys support was compiled.
4447 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4449 PH/24 Added ${if match_ip condition.
4451 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4452 databases so that it will be absolutely obvious if a crash occurs in the
4453 DB library. This is a regular occurrence (often caused by mis-matched
4456 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4457 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4458 picked up on a user's system that detects such things. There doesn't seem
4459 to be a gcc warning option for this - only an attribute that has to be
4460 put on the function's prototype. It seems that in Fedora Core 4 they have
4461 set this on a number of new functions. No doubt there will be more in due
4464 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4465 specify a transport. However, if an address that was verified by such a
4466 router was the subject of a callout, Exim crashed because it tried to
4467 read the rcpt_include_affixes from the non-existent transport. Now it
4468 just assumes that the setting of that option is false. This bug was
4469 introduced by 4.51/PH/31.
4471 PH/28 Changed -d+all to exclude +memory, because that information is very
4472 rarely of interest, but it makes the output a lot bigger. People tend to
4473 do -d+all out of habit.
4475 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4476 code in os-type was giving problems when libc.so lives in lib64, like on
4479 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4480 aren't the modern standard, and it seems that some systems' include files
4481 don't always have them. Exim was already checking for some of the newer
4482 ones like T_AAAA, and defining it itself. I've added checks for all the
4483 record types that Exim uses.
4485 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4486 not automatically generating a new one, as it is supposed to. This
4487 prevented TLS from working. If the file did exist, but contained invalid
4488 data, a new version was generated, as expected. It was only the case of a
4489 non-existent file that was broken.
4491 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4492 with a change in libdomainkeys > 0.64.
4494 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4495 from DNS. If the selector record carries the flag, it now has
4496 precedence over the domain-wide flag.
4498 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4500 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4501 the use of a port name that isn't defined in /etc/services) occurred, the
4502 message was deferred as in a normal delivery, and thus remained on the
4503 spool, instead of being failed because of the mua_wrapper setting. This
4504 is now fixed, and I tidied up some of the mua_wrapper messages at the
4507 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4508 the messages in a hash of arrays rather than using individual hashes.
4509 This is a bit cleaner and results in dramatic memory savings, albeit
4510 at a slight CPU cost.
4512 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4513 as requested by Marc Sherman.
4515 SC/10 Eximstats: added histograms for user specified patterns as requested
4518 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4520 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4521 fopen() in the content-scanning modules that did not already have it.
4527 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4528 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4530 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4532 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4534 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4535 to test Sieve filters that use "vacation".
4537 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4538 that changes the way the GnuTLS parameters are stored in the cache file.
4539 The new format can be generated externally. For backward compatibility,
4540 if the data in the cache doesn't make sense, Exim assumes it has read an
4541 old-format file, and it generates new data and writes a new file. This
4542 means that you can't go back to an older release without removing the
4545 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4546 work if there are any delivery delays because "one_time" forces the
4547 parent to be marked "delivered", so its unseen clone is never tried
4548 again. For this reason, Exim now forbids the simultaneous setting of
4551 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4552 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4553 entry if you want to know the details. The fix, however, neglected to
4554 consider the case when local delivery batching is involved. The test for
4555 "previously delivered" was not happening when checking to see if an
4556 address could be batched with a previous (undelivered) one; under
4557 certain circumstances this could lead to multiple deliveries to the same
4560 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4561 in its include files, and this causes problems building Exim.
4563 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4564 header_syntax) but Exim was just ignoring anything given after a slash.
4565 In particular, this caused confusion with an attempt to use "verify =
4566 reverse_host_lookup/defer_ok". An error is now given when options are
4567 supplied for verify items that do not have them. (Maybe reverse_host_
4568 lookup should have a defer_ok option, but that's a different point.)
4570 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4571 defined by RFC 821) to 2048, because there were problems with some AUTH
4572 commands, and RFC 1869 says the size should be increased for extended
4573 SMTP commands that take arguments.
4575 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4578 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4579 "unknown" error; now it says that the functionality isn't in the binary.
4581 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4582 an address' error message when a string expansion fails (syntax or
4583 whatever). Otherwise the password may appear in the log. Following change
4584 PH/42 below, there is no longer a chance of it appearing in a bounce
4587 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4589 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4590 of its addresses was ever tried. (Bugzilla bug #2.)
4592 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4593 the result incorrectly in the debug output. (It correctly added a newline
4594 to what was transported.)
4596 TF/01 Added $received_time.
4598 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4599 commented out examples of how to interface to a virus scanner and to
4600 SpamAssassin. Also added commented examples of av_scanner and
4601 spamd_address settings.
4603 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4604 and controls are allowed in which ACLs. There were a couple of minor
4605 errors. Some of the entries in the conditions table (which is a table of
4606 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4607 negation of where the condition IS allowed.
4609 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4611 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4612 header file does not have a version number, so I've had to invent a new
4613 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4614 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4615 radiusclient), but it was contributed by a Radius user.
4617 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4618 files or directories, for OpenSSL.
4620 PH/19 When an Exim process that is running as root has to create an Exim log
4621 file, it does so in a subprocess that runs as exim:exim so as to get the
4622 ownership right at creation (otherwise, other Exim processes might see
4623 the file with the wrong ownership). There was no test for failure of this
4624 fork() call, which would lead to the process getting stuck as it waited
4625 for a non-existent subprocess. Forks do occasionally fail when resources
4626 run out. I reviewed all the other calls to fork(); they all seem to check
4629 PH/20 When checking for unexpected SMTP input at connect time (before writing
4630 the banner), Exim was not dealing correctly with a non-positive return
4631 from the read() function. If the client had disconnected by this time,
4632 the result was a log entry for a synchronization error with an empty
4633 string after "input=" when read() returned zero. If read() returned -1
4634 (an event I could not check), uninitialized data bytes were printed.
4635 There were reports of junk text (parts of files, etc) appearing after
4638 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4640 PH/22 Added support for macro redefinition, and (re)definition in between
4641 driver and ACL definitions.
4643 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4644 forgetting to use the resulting value; it was using the unexpanded value.
4646 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4647 hadn't been configured. The fix is from Juergen Kreileder, who
4648 understands it better than I do:
4650 "Here's what I see happening with three configured cyrus_sasl
4651 authenticators configured (plain, login, cram-md5):
4653 On startup auth_cyrus_sasl_init() gets called for each of these.
4654 This means three calls to sasl_listmech() without a specified mech_list.
4655 => SASL tests which mechs of all available mechs actually work
4656 => three warnings about OTP not working
4657 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4659 With the patch, sasl_listmech() also gets called three times. But now
4660 SASL's mech_list option is set to the server_mech specified in the the
4661 authenticator. Or in other words, the answer from sasl_listmech()
4662 gets limited to just the mech you're testing for (which is different
4664 => the return list contains just 'plain' or 'login', 'cram-md5' or
4665 nothing depending on the value of ob->server_mech.
4667 I've just tested the patch: Authentication still works fine,
4668 unavailable mechs specified in the exim configuration are still
4669 caught, and the auth.log warnings about OTP are gone."
4671 PH/25 When debugging is enabled, the contents of the command line are added
4672 to the debugging output, even when log_selector=+arguments is not
4675 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4676 answer is "GNU", and only if the return is "GNU/something" is the answer
4679 PH/27 $acl_verify_message is now set immediately after the failure of a
4680 verification in an ACL, and so is available in subsequent modifiers. In
4681 particular, the message can be preserved by coding like this:
4683 warn !verify = sender
4684 set acl_m0 = $acl_verify_message
4686 Previously, $acl_verify_message was set only while expanding "message"
4687 and "log_message" when a very denied access.
4689 PH/28 Modified OS/os.c-Linux with
4691 -#ifndef OS_LOAD_AVERAGE
4692 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4694 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4695 nomenclature these days.)
4697 PH/29 Installed patch from the Sieve maintainer that adds the options
4698 sieve_useraddress and sieve_subaddress to the redirect router.
4700 PH/30 In these circumstances:
4701 . Two addresses routed to the same list of hosts;
4702 . First host does not offer TLS;
4703 . First host accepts first address;
4704 . First host gives temporary error to second address;
4705 . Second host offers TLS and a TLS session is established;
4706 . Second host accepts second address.
4707 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4708 and peerdn, if requested) that were in fact used only for the second
4711 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4712 attention to any local part prefix or suffix that was matched by the
4713 router that accepted the address. It now behaves in the same way as it
4714 does for delivery: the affixes are removed from the local part unless
4715 rcpt_include_affixes is set on the transport.
4717 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4718 timeout during the DATA phase of an incoming message.
4720 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4721 applied a patch sanctioned by the Sieve maintainer.
4723 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4724 the uid or gid is negative. A case of a negative gid caused this to be
4725 noticed. The fix allows for either to be negative.
4727 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4728 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4731 PH/36 The change PH/12 above was broken. Fixed it.
4733 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4734 the grounds that routing the same address twice would always produce the
4735 same answer. This might have been true once, but it is certainly no
4736 longer true now. Routing a child address may depend on the previous
4737 routing that produced that child. Some complicated redirection strategies
4738 went wrong when messages had multiple recipients, and made Exim's
4739 behaviour dependent on the order in which the addresses were given.
4741 I have moved the duplicate checking until after the routing is complete.
4742 Exim scans the addresses that are assigned to local and remote
4743 transports, and removes any duplicates. This means that more work will be
4744 done, as duplicates will always all be routed, but duplicates are
4745 presumably rare, so I don't expect this is of any significance.
4747 For deliveries to pipes, files, and autoreplies, the duplicate checking
4748 still happens during the routing process, since they are not going to be
4751 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4752 It corrects a timeout issue with spamd. This is Ian's comment: "The
4753 background is that sometimes spamd either never reads data from a
4754 connection it has accepted, or it never writes response data. The exiscan
4755 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4756 blindly assumes that writes won't block so it may never time out."
4758 PH/39 Allow G after quota size as well as K and M.
4760 PH/40 The value set for $authenticated_id in an authenticator may not contain
4761 binary zeroes or newlines because the value is written to log lines and
4762 to spool files. There was no check on this. Now the value is run through
4763 the string_printing() function so that such characters are converted to
4764 printable escape sequences.
4766 PH/41 $message_linecount is a new variable that contains the total number of
4767 lines in the message. Compare $body_linecount, which is the count for the
4770 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4771 bounce and delay warning messages, except in certain special cases, which
4774 (a) An SMTP error message from a remote host;
4775 (b) A message specified in a :fail: redirection;
4776 (c) A message specified in a "fail" command in a system filter;
4777 (d) A message specified in a FAIL return from the queryprogram router;
4778 (e) A message specified by the cannot_route_message router option.
4780 In these cases only, Exim does include the error details in bounce and
4781 warning messages. There are also a few cases where bland messages such
4782 as "unrouteable address" or "local delivery error" are given.
4784 PH/43 $value is now also set for the "else" part of a ${run expansion.
4786 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4787 being worked on, but at least Exim now implements the latest version to
4790 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4791 process to complete was treated as a delivery failure, a timeout while
4792 writing the message to the pipe was logged, but erroneously treated as a
4793 successful delivery. Such timeouts include transport filter timeouts. For
4794 consistency with the overall process timeout, these timeouts are now
4795 treated as errors, giving rise to delivery failures by default. However,
4796 there is now a new Boolean option for the pipe transport called
4797 timeout_defer, which, if set TRUE, converts the failures into defers for
4798 both kinds of timeout. A transport filter timeout is now identified in
4801 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4802 systems where "make" and "gmake" are different, calling "gmake" at top
4803 level broke things. I've arranged for the value of $(MAKE) to be passed
4804 from the Makefile to this script so that it can call the same version of
4808 A note about Exim versions 4.44 and 4.50
4809 ----------------------------------------
4811 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4812 changes of various kinds. As a consequence, a big documentation update was
4813 needed. This delayed the release for rather longer than seemed good, especially
4814 in the light of a couple of (minor) security issues. Therefore, the changes
4815 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4816 release. So 4.44 and 4.50 are in effect two different branches that both start
4819 I have left the 4.50 change log unchanged; it contains all the changes since
4820 4.43. The change log for 4.44 is below; many of its items are identical to
4821 those for 4.50. This seems to be the most sensible way to preserve the
4822 historical information.
4828 1. Minor wording change to the doc/README.SIEVE file.
4830 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4831 computation of the current number of files was incorrect.
4833 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4834 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4837 4. Give more explanation in the error message when the command for a transport
4838 filter fails to execute.
4840 5. There are several places where Exim runs a non-Exim command in a
4841 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4842 was being done only for the command run by the queryprogram router. It is
4843 now done for all such subprocesses. The other cases are: ${run, transport
4844 filters, and the commands run by the lmtp and pipe transports.
4846 6. Added CONFIGURE_GROUP build-time option.
4848 7. Some older OS have a limit of 256 on the maximum number of file
4849 descriptors. Exim was using setrlimit() to set 1000 as a large value
4850 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4851 systems. I've change it so that if it can't get 1000, it tries for 256.
4853 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4854 was an oversight, and furthermore, ever since the addition of extra
4855 controls (e.g. 4.43/32), the checks on when to allow different forms of
4856 "control" were broken. There should now be diagnostics for all cases when a
4857 control that does not make sense is encountered.
4859 9. Added the /retain_sender option to "control=submission".
4861 10. $recipients is now available in the predata ACL (oversight).
4863 11. Tidy the search cache before the fork to do a delivery from a message
4864 received from the command line. Otherwise the child will trigger a lookup
4865 failure and thereby defer the delivery if it tries to use (for example) a
4866 cached ldap connection that the parent has called unbind on.
4868 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4869 of $address_data from the recipient verification was clobbered by the
4870 sender verification.
4872 13. The value of address_data from a sender verification is now available in
4873 $sender_address_data in subsequent conditions in the ACL statement.
4875 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4877 15. Added a new option "connect=<time>" to callout options, to set a different
4880 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4881 was its contents. (It was OK if the option was not defined at all.)
4883 17. A "Completed" log line is now written for messages that are removed from
4884 the spool by the -Mrm option.
4886 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4887 information about exactly what failed.
4889 19. Added -dd to debug only the daemon process.
4891 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4892 handles timeouts, both on the server side and network timeouts. Renamed the
4893 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4895 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4896 It is now set to "smtps".
4898 22. $host_address is now set to the target address during the checking of
4899 ignore_target_hosts.
4901 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4902 being passed; this would have caused $sender_host_name to have been used if
4903 matching the list had actually called for a host name (not very likely,
4904 since this list is usually IP addresses). A host name is now passed as
4907 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4908 code that specifies a non-restarting handler (typically sigaction() in
4909 modern systems) in an attempt to fix a rare and obscure crash bug.
4911 25. Narrowed the window for a race in the daemon that could cause it to ignore
4912 SIGCHLD signals. This is not a major problem, because they are used only to
4913 wake it up if nothing else does.
4915 26. A malformed maildirsize file could cause Exim to calculate negative values
4916 for the mailbox size or file count. Odd effects could occur as a result.
4917 The maildirsize information is now recalculated if the size or filecount
4920 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4921 support for a long time. Removed HAVE_SYS_VFS_H.
4923 28. Installed the latest version of exipick from John Jetmore.
4925 29. In an address list, if the pattern was not a regular expression, an empty
4926 subject address (from a bounce message) matched only if the pattern was an
4927 empty string. Non-empty patterns were not even tested. This was the wrong
4928 because it is perfectly reasonable to use an empty address as part of a
4929 database query. An empty address is now tested by patterns that are
4930 lookups. However, all the other forms of pattern expect the subject to
4931 contain a local part and a domain, and therefore, for them, an empty
4932 address still always fails if the pattern is not itself empty.
4934 30. Exim went into a mad DNS loop when attempting to do a callout where the
4935 host was specified on an smtp transport, and looking it up yielded more
4936 than one IP address.
4938 31. Re-factored the code for checking spool and log partition space into a
4939 function that finds that data and another that does the check. The former
4940 is then used to implement four new variables: $spool_space, $log_space,
4941 $spool_inodes, and $log_inodes.
4943 32. The RFC2047 encoding function was originally intended for short strings
4944 such as real names; it was not keeping to the 75-character limit for
4945 encoded words that the RFC imposes. It now respects the limit, and
4946 generates multiple encoded words if necessary. To be on the safe side, I
4947 have increased the buffer size for the ${rfc2047: expansion operator from
4950 33. It is now permitted to omit both strings after an "if" condition; if the
4951 condition is true, the result is "true". As before, when the second string
4952 is omitted, a false condition yields an empty string. This makes it less
4953 cumbersome to write custom ACL and router conditions.
4955 34. Failure to deliver a bounce message always caused it to be frozen, even if
4956 there was an errors_to setting on the router. The errors_to setting is now
4959 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4960 canonical form (fully expanded) before being placed in
4961 $sender_host_address.
4963 36. The table in the code that translates DNS record types into text (T_A to
4964 "A" for instance) was missing entries for NS and CNAME. It is just possible
4965 that this could have caused confusion if both these types were looked up
4966 for the same domain, because the text type is used as part of Exim's
4967 per-process caching. But the chance of anyone hitting this buglet seems
4970 37. The dnsdb lookup has been extended in a number of ways.
4972 (1) There is a new type, "zns", which walks up the domain tree until it
4973 finds some nameserver records. It should be used with care.
4975 (2) There is a new type, "mxh", which is like "mx" except that it returns
4976 just the host names, not the priorities.
4978 (3) It is now possible to give a list of domains (or IP addresses) to be
4979 looked up. The behaviour when one of the lookups defers can be
4980 controlled by a keyword.
4982 (4) It is now possible to specify the separator character for use when
4983 multiple records are returned.
4985 38. The dnslists ACL condition has been extended: it is now possible to supply
4986 a list of IP addresses and/or domains to be looked up in a particular DNS
4989 39. Added log_selector=+queue_time_overall.
4991 40. When running the queue in the test harness, wait just a tad after forking a
4992 delivery process, to get repeatability of debugging output.
4994 41. Include certificate and key file names in error message when GnuTLS fails
4995 to set them up, because the GnuTLS error message doesn't include the name
4996 of the failing file when there is a problem reading it.
4998 42. Allow both -bf and -bF in the same test run.
5000 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5002 44. The "Exiscan patch" is now merged into the mainline Exim source.
5004 45. Sometimes the final signoff response after QUIT could fail to get
5005 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5006 < 0 before doing a fflush(). This bug looks as though it goes back to the
5007 introduction of TLS in release 3.20, but "sometimes" must have been rare
5008 because the tests only now provoked it.
5010 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5011 (this can affect the format of dates).
5013 47. exim_tidydb, when checking for the continued existence of a message for
5014 which it has found a message-specific retry record, was not finding
5015 messages that were in split spool directories. Consequently, it was
5016 deleting retry records that should have stayed in existence.
5018 48. Steve fixed some bugs in eximstats.
5020 49. The SPA authentication driver was not abandoning authentication and moving
5021 on to the next authenticator when an expansion was forced to fail,
5022 contradicting the general specification for all authenticators. Instead it
5023 was generating a temporary error. It now behaves as specified.
5025 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5026 (the order specifies the preference for clients). The order is now AES256,
5027 AES128, 3DES, ARCFOUR128.
5029 51. Small patch to Sieve code - explicitly set From: when generating an
5032 52. Exim crashed if a remote delivery caused a very long error message to be
5033 recorded - for instance if somebody sent an entire SpamAssassin report back
5034 as a large number of 550 error lines. This bug was coincidentally fixed by
5035 increasing the size of one of Exim's internal buffers (big_buffer) that
5036 happened as part of the Exiscan merge. However, to be on the safe side, I
5037 have made the code more robust (and fixed the comments that describe what
5040 53. Now that there can be additional text after "Completed" in log lines (if
5041 the queue_time_overall log selector is set), a one-byte patch to exigrep
5042 was needed to allow it to recognize "Completed" as not the last thing in
5045 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5046 patch that reportedly fixes this has been added. I am not expert enough to
5047 create a test for it. This is what the patch creator wrote:
5049 "I found a little strange behaviour of ldap code when working with
5050 Windows 2003 AD Domain, where users was placed in more than one
5051 Organization Units. When I tried to give exim partial DN, the exit code
5052 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5053 But simultaneously result of request was absolutely normal ldap result,
5054 so I produce this patch..."
5056 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5057 REFERENCE, so I have modified the code to exclude the patch when that macro
5060 55. Some experimental protocols are using DNS PTR records for new purposes. The
5061 keys for these records are domain names, not reversed IP addresses. The
5062 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5063 leaves it alone. Component reversal etc. now happens only for IP addresses.
5066 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5068 57. Double the size of the debug message buffer (to 2048) so that more of very
5069 long debug lines gets shown.
5071 58. The exicyclog utility now does better if the number of log files to keep
5072 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5074 59. Two changes related to the smtp_active_hostname option:
5076 (1) $smtp_active_hostname is now available as a variable.
5077 (2) The default for smtp_banner uses $smtp_active_hostname instead
5078 of $primary_hostname.
5080 60. The host_aton() function is supposed to be passed a string that is known
5081 to be a valid IP address. However, in the case of IPv6 addresses, it was
5082 not checking this. This is a hostage to fortune. Exim now panics and dies
5083 if the condition is not met. A case was found where this could be provoked
5084 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5085 components; fortuitously, this particular loophole had already been fixed
5086 by change 4.50/55 above.
5088 If there are any other similar loopholes, the new check in host_aton()
5089 itself should stop them being exploited. The report I received stated that
5090 data on the command line could provoke the exploit when Exim was running as
5091 exim, but did not say which command line option was involved. All I could
5092 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5093 running as the user.
5096 61. There was a buffer overflow vulnerability in the SPA authentication code
5097 (which came originally from the Samba project). I have added a test to the
5098 spa_base64_to_bits() function which I hope fixes it.
5101 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5102 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5104 63. The daemon start-up calls getloadavg() while still root for those OS that
5105 need the first call to be done as root, but it missed one case: when
5106 deliver_queue_load_max is set with deliver_drop_privilege. This is
5107 necessary for the benefit of the queue runner, because there is no re-exec
5108 when deliver_drop_privilege is set.
5110 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5111 This has been fixed.
5113 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5114 was in use, was not putting the data itself into the right store pool;
5115 consequently, it could be overwritten for a subsequent message in the same
5116 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5119 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5121 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5122 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5123 the function were treating the return as a boolean value, which happened to
5124 work because 0=false and not-0=true, but is not correct code.
5126 68. The host_aton() function was not handling scoped IPv6 addresses (those
5127 with, for example, "%eth0" on the end) correctly.
5129 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5130 negated items (that is, ~something) in unsigned ints. Some compilers
5131 apparently mutter when there is no cast.
5133 70. If an address verification called from an ACL failed, and did not produce a
5134 user-specific message (i.e. there was only a "system" message), nothing was
5135 put in $acl_verify_message. In this situation, it now puts the system
5138 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5139 session; change 4.31/43 added the unwanted input to the log line - except
5140 that it did not do this in the start of session case. It now does.
5142 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5143 This is wrong and can cause the other end to generate a synchronization
5144 error if it is another Exim or anything else that does the synchronization
5145 check. A QUIT command is no longer sent after a timeout.
5147 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5148 during host lookups.
5150 74. The defer_ok option of callout verification was not working if it was used
5151 when verifying addresses in header lines, that is, for this case:
5153 verify = header_sender/callout=defer_ok
5155 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5156 those file descriptors could be used for SMTP connections. If anything
5157 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5158 could be sent to the SMTP client, causing chaos. The daemon now opens
5159 stdin, stdout, and stderr to /dev/null when it puts itself into the
5162 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5163 log by default. The user can override this with suitable Perl magic.
5165 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5166 the log message when discard triggers, was not working for the DATA ACL or
5167 for the non-SMTP ACL.
5169 78. Error message wording change in sieve.c.
5171 79. If smtp_accept_max_per_host was set, the number of connections could be
5172 restricted to fewer than expected, because the daemon was trying to set up
5173 a new connection before checking whether the processes handling previous
5174 connections had finished. The check for completed processes is now done
5175 earlier. On busy systems, this bug wouldn't be noticed because something
5176 else would have woken the daemon, and it would have reaped the completed
5179 80. If a message was submitted locally by a user whose login name contained one
5180 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5181 It caused a spool format error. I have fixed the spool reading code. A
5182 related problem was that the "from" clause in the Received: line became
5183 illegal because of the space(s). It is now covered by ${quote_local_part.
5185 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5188 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5190 83. Patch from Sieve maintainer to fix unterminated string problem in
5191 "vacation" handling.
5193 84. Some minor changes to the Linux configuration files to help with other
5194 OS variants using glibc.
5196 85. One more patch for Sieve to update vacation handling to latest spec.
5199 ----------------------------------------------------
5200 See the note above about the 4.44 and 4.50 releases.
5201 ----------------------------------------------------
5207 1. Change 4.43/35 introduced a bug that caused file counts to be
5208 incorrectly computed when quota_filecount was set in an appendfile
5211 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5212 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5215 3. Give more explanation in the error message when the command for a transport
5216 filter fails to execute.
5218 4. There are several places where Exim runs a non-Exim command in a
5219 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5220 was being done only for the command run by the queryprogram router. It is
5221 now done for all such subprocesses. The other cases are: ${run, transport
5222 filters, and the commands run by the lmtp and pipe transports.
5224 5. Some older OS have a limit of 256 on the maximum number of file
5225 descriptors. Exim was using setrlimit() to set 1000 as a large value
5226 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5227 systems. I've change it so that if it can't get 1000, it tries for 256.
5229 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5230 was an oversight, and furthermore, ever since the addition of extra
5231 controls (e.g. 4.43/32), the checks on when to allow different forms of
5232 "control" were broken. There should now be diagnostics for all cases when a
5233 control that does not make sense is encountered.
5235 7. $recipients is now available in the predata ACL (oversight).
5237 8. Tidy the search cache before the fork to do a delivery from a message
5238 received from the command line. Otherwise the child will trigger a lookup
5239 failure and thereby defer the delivery if it tries to use (for example) a
5240 cached ldap connection that the parent has called unbind on.
5242 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5243 of $address_data from the recipient verification was clobbered by the
5244 sender verification.
5246 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5247 was its contents. (It was OK if the option was not defined at all.)
5249 11. A "Completed" log line is now written for messages that are removed from
5250 the spool by the -Mrm option.
5252 12. $host_address is now set to the target address during the checking of
5253 ignore_target_hosts.
5255 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5256 being passed; this would have caused $sender_host_name to have been used if
5257 matching the list had actually called for a host name (not very likely,
5258 since this list is usually IP addresses). A host name is now passed as
5261 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5262 code that specifies a non-restarting handler (typically sigaction() in
5263 modern systems) in an attempt to fix a rare and obscure crash bug.
5265 15. Narrowed the window for a race in the daemon that could cause it to ignore
5266 SIGCHLD signals. This is not a major problem, because they are used only to
5267 wake it up if nothing else does.
5269 16. A malformed maildirsize file could cause Exim to calculate negative values
5270 for the mailbox size or file count. Odd effects could occur as a result.
5271 The maildirsize information is now recalculated if the size or filecount
5274 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5275 support for a long time. Removed HAVE_SYS_VFS_H.
5277 18. Updated exipick to current release from John Jetmore.
5279 19. Allow an empty sender to be matched against a lookup in an address list.
5280 Previously the only cases considered were a regular expression, or an
5283 20. Exim went into a mad DNS lookup loop when doing a callout where the
5284 host was specified on the transport, if the DNS lookup yielded more than
5287 21. The RFC2047 encoding function was originally intended for short strings
5288 such as real names; it was not keeping to the 75-character limit for
5289 encoded words that the RFC imposes. It now respects the limit, and
5290 generates multiple encoded words if necessary. To be on the safe side, I
5291 have increased the buffer size for the ${rfc2047: expansion operator from
5294 22. Failure to deliver a bounce message always caused it to be frozen, even if
5295 there was an errors_to setting on the router. The errors_to setting is now
5298 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5299 canonical form (fully expanded) before being placed in
5300 $sender_host_address.
5302 24. Updated eximstats to version 1.33
5304 25. Include certificate and key file names in error message when GnuTLS fails
5305 to set them up, because the GnuTLS error message doesn't include the name
5306 of the failing file when there is a problem reading it.
5308 26. Expand error message when OpenSSL has problems setting up cert/key files.
5311 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5312 (this can affect the format of dates).
5314 28. exim_tidydb, when checking for the continued existence of a message for
5315 which it has found a message-specific retry record, was not finding
5316 messages that were in split spool directories. Consequently, it was
5317 deleting retry records that should have stayed in existence.
5319 29. eximstats updated to version 1.35
5320 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5321 1.35 - bugfix such that pie charts by volume are generated correctly
5323 30. The SPA authentication driver was not abandoning authentication and moving
5324 on to the next authenticator when an expansion was forced to fail,
5325 contradicting the general specification for all authenticators. Instead it
5326 was generating a temporary error. It now behaves as specified.
5328 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5329 (the order specifies the preference for clients). The order is now AES256,
5330 AES128, 3DES, ARCFOUR128.
5332 31. Small patch to Sieve code - explicitly set From: when generating an
5335 32. Exim crashed if a remote delivery caused a very long error message to be
5336 recorded - for instance if somebody sent an entire SpamAssassin report back
5337 as a large number of 550 error lines. This bug was coincidentally fixed by
5338 increasing the size of one of Exim's internal buffers (big_buffer) that
5339 happened as part of the Exiscan merge. However, to be on the safe side, I
5340 have made the code more robust (and fixed the comments that describe what
5343 33. Some experimental protocols are using DNS PTR records for new purposes. The
5344 keys for these records are domain names, not reversed IP addresses. The
5345 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5346 leaves it alone. Component reversal etc. now happens only for IP addresses.
5349 34. The host_aton() function is supposed to be passed a string that is known
5350 to be a valid IP address. However, in the case of IPv6 addresses, it was
5351 not checking this. This is a hostage to fortune. Exim now panics and dies
5352 if the condition is not met. A case was found where this could be provoked
5353 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5354 components; fortuitously, this particular loophole had already been fixed
5355 by change 4.50/55 or 4.44/33 above.
5357 If there are any other similar loopholes, the new check in host_aton()
5358 itself should stop them being exploited. The report I received stated that
5359 data on the command line could provoke the exploit when Exim was running as
5360 exim, but did not say which command line option was involved. All I could
5361 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5362 running as the user.
5365 35. There was a buffer overflow vulnerability in the SPA authentication code
5366 (which came originally from the Samba project). I have added a test to the
5367 spa_base64_to_bits() function which I hope fixes it.
5370 36. The daemon start-up calls getloadavg() while still root for those OS that
5371 need the first call to be done as root, but it missed one case: when
5372 deliver_queue_load_max is set with deliver_drop_privilege. This is
5373 necessary for the benefit of the queue runner, because there is no re-exec
5374 when deliver_drop_privilege is set.
5376 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5377 was in use, was not putting the data itself into the right store pool;
5378 consequently, it could be overwritten for a subsequent message in the same
5379 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5382 38. Sometimes the final signoff response after QUIT could fail to get
5383 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5384 < 0 before doing a fflush(). This bug looks as though it goes back to the
5385 introduction of TLS in release 3.20, but "sometimes" must have been rare
5386 because the tests only now provoked it.
5392 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5393 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5394 Now it returns an integer. A number of calls were still expecting a T/F
5395 return. Fortuitously, in all cases, the tests worked in OK situations,
5396 which is the norm. However, things would have gone wrong on any write
5397 failures on the smtp file descriptor. This function is used when sending
5398 messages over SMTP and also when doing verify callouts.
5400 2. When Exim is called to do synchronous delivery of a locally submitted
5401 message (the -odf or -odi options), it no longer closes stderr before doing
5404 3. Implemented the mua_wrapper option.
5406 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5408 5. Implemented the functions header_remove(), header_testname(),
5409 header_add_at_position(), and receive_remove_recipient(), and exported them
5412 6. If an ACL "warn" statement specified the addition of headers, Exim already
5413 inserted X-ACL-Warn: at the start if there was no header name. However, it
5414 was not making this test for the second and subsequent header lines if
5415 there were newlines in the string. This meant that an invalid header could
5416 be inserted if Exim was badly configured.
5418 7. Allow an ACL "warn" statement to add header lines at the start or after all
5419 the Received: headers, as well as at the end.
5421 8. Added the rcpt_4xx retry error code.
5423 9. Added postmaster_mailfrom=xxx to callout verification option.
5425 10. Added mailfrom=xxxx to the callout verification option, for verify=
5428 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5429 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5431 12. Inserted some casts to stop certain compilers warning when using pointer
5432 differences as field lengths or precisions in printf-type calls (mostly
5433 affecting debugging statements).
5435 13. Added optional readline() support for -be (dynamically loaded).
5437 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5438 same clock tick as a message's arrival, so that its received time was the
5439 same as the "first fail" time on the retry record, and that message
5440 remained on the queue past the ultimate address timeout, every queue runner
5441 would try a delivery (because it was past the ultimate address timeout) but
5442 after another failure, the ultimate address timeout, which should have then
5443 bounced the address, did not kick in. This was a "< instead of <=" error;
5444 in most cases the first failure would have been in the next clock tick
5445 after the received time, and all would be well.
5447 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5448 being recognized when the domain list was tested by the match_domain
5449 condition in an expansion string.
5451 16. Added the ${str2b64: operator.
5453 17. Exim was always calling setrlimit() to set a large limit for the number of
5454 processes, without checking whether the existing limit was already
5455 adequate. (It did check for the limit on file descriptors.) Furthermore,
5456 errors from getrlimit() and setrlimit() were being ignored. Now they are
5457 logged to the main and panic logs, but Exim does carry on, to try to do its
5458 job under whatever limits there are.
5460 18. Imported PCRE 5.0.
5462 19. Trivial typo in log message " temporarily refused connection" (the leading
5465 20. If the log selector return_path_on_delivery was set and an address was
5466 redirected to /dev/null, the delivery process crashed because it assumed
5467 that a return path would always be set for a "successful" delivery. In this
5468 case, the whole delivery is bypassed as an optimization, and therefore no
5471 21. Internal re-arrangement: the function for sending a challenge and reading
5472 a response while authentication was assuming a zero-terminated challenge
5473 string. It's now changed to take a pointer and a length, to allow for
5474 binary data in such strings.
5476 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5478 23. Exim was not respecting finduser_retries when seeking the login of the
5479 uid under which it was called; it was always trying 10 times. (The default
5480 setting of finduser_retries is zero.) Also, it was sleeping after the final
5481 failure, which is pointless.
5483 24. Implemented tls_on_connect_ports.
5485 25. Implemented acl_smtp_predata.
5487 26. If the domain in control=submission is set empty, Exim assumes that the
5488 authenticated id is a complete email address when it generates From: or
5489 Sender: header lines.
5491 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5492 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5493 chown and chgrp in /bin and hostname in /usr/bin.
5495 28. Exim was keeping the "process log" file open after each use, just as it
5496 does for the main log. This opens the possibility of it remaining open for
5497 long periods when the USR1 signal hits a daemon. Occasional processlog
5498 errors were reported, that could have been caused by this. Anyway, it seems
5499 much more sensible not to leave this file open at all, so that is what now
5502 29. The long-running daemon process does not normally write to the log once it
5503 has entered its main loop, and it closes the log before doing so. This is
5504 so that log files can straightforwardly be renamed and moved. However,
5505 there are a couple of unusual error situations where the daemon does write
5506 log entries, and I had neglected to close the log afterwards.
5508 30. The text of an SMTP error response that was received during a remote
5509 delivery was being truncated at 512 bytes. This is too short for some of
5510 the long messages that one sometimes sees. I've increased the limit to
5513 31. It is now possible to make retry rules that apply only when a message has a
5514 specific sender, in particular, an empty sender.
5516 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5517 it possible to be selective about when SMTP synchronization is enforced.
5519 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5521 32. Implemented hosts_connection_nolog.
5523 33. Added an ACL for QUIT.
5525 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5528 35. Added mailbox_size and mailbox_filecount to appendfile.
5530 36. Added control = no_multiline_responses to ACLs.
5532 37. There was a bug in the logic of the code that waits for the clock to tick
5533 in the case where the clock went backwards by a substantial amount such
5534 that the microsecond fraction of "now" was more than the microsecond
5535 fraction of "then" (but the whole seconds number was less).
5537 38. Added support for the libradius Radius client library this is found on
5538 FreeBSD (previously only the radiusclient library was supported).
5544 1. When certain lookups returned multiple values in the form name=value, the
5545 quoting of the values was not always being done properly. Specifically:
5546 (a) If the value started with a double quote, but contained no whitespace,
5548 (b) If the value contained whitespace other than a space character (i.e.
5549 tabs or newlines or carriage returns) it was not quoted.
5550 This fix has been applied to the mysql and pgsql lookups by writing a
5551 separate quoting function and calling it from the lookup code. The fix
5552 should probably also be applied to nisplus, ibase and oracle lookups, but
5553 since I cannot test any of those, I have not disturbed their existing code.
5555 2. A hit in the callout cache for a specific address caused a log line with no
5556 reason for rejecting RCPT. Now it says "Previous (cached) callout
5557 verification failure".
5559 3. There was an off-by-one bug in the queryprogram router. An over-long
5560 return line was truncated at 256 instead of 255 characters, thereby
5561 overflowing its buffer with the terminating zero. As well as fixing this, I
5562 have increased the buffer size to 1024 (and made a note to document this).
5564 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5565 when Exim is waiting for an SMTP response from a remote server, Exim
5566 restarts its select() call on the socket, thereby resetting its timeout.
5567 This is not a problem when such interrupts are rare. Somebody set up a cron
5568 job to run exiwhat every 2 minutes, which is less than the normal select()
5569 timeout (5 or 10 minutes). This meant that the select() timeout never
5570 kicked in because it was always reset. I have fixed this by comparing the
5571 time when an interrupt arrives with the time at the start of the first call
5572 to select(). If more time than the timeout has elapsed, the interrupt is
5573 treated as a timeout.
5575 5. Some internal re-factoring in preparation for the addition of Sieve
5576 extensions (by MH). In particular, the "personal" test is moved to a
5577 separate function, and given an option for scanning Cc: and Bcc: (which is
5578 not set for Exim filters).
5580 6. When Exim created an email address using the login of the caller as the
5581 local part (e.g. when creating a From: or Sender: header line), it was not
5582 quoting the local part when it contained special characters such as @.
5584 7. Installed new OpenBSD configuration files.
5586 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5587 try to make them clearer.
5589 9. Callout options, other than the timeout value, were being ignored when
5590 verifying sender addresses in header lines. For example, when using
5592 verify = header_sender/callout=no_cache
5594 the cache was (incorrectly) being used.
5596 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5597 only the Cygwin environment.
5599 11. When return_path_on_delivery was set as a log selector, if different remote
5600 addresses in the same message used different return paths and parallel
5601 remote delivery occurred, the wrong values would sometimes be logged.
5602 (Whenever a remote delivery process finished, the return path value from
5603 the most recently started remote delivery process was logged.)
5605 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5606 lines when AUTH and/or TLS are in use. This is the "received protocol"
5607 field. Exim used to use "asmtp" for authenticated SMTP, without any
5608 indication (in the protocol name) for TLS use. Now it follows the RFC and
5609 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5610 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5611 names appear in log lines as well as in Received: header lines.
5613 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5614 extensions, and comparison tests, and to fix some bugs.
5616 14. Changes to the "personal" filter test:
5618 (1) The test was buggy in that it was just doing the equivalent of
5619 "contains" tests on header lines. For example, if a user's address was
5620 anne@some.where, the "personal" test would incorrectly be true for
5622 To: susanne@some.where
5624 This test is now done by extracting each address from the header in turn,
5625 and checking the entire address. Other tests that are part of "personal"
5626 are now done using regular expressions (for example, to check local parts
5627 of addresses in From: header lines).
5629 (2) The list of non-personal local parts in From: addresses has been
5630 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5631 taken from the Sieve specification recommendations.
5633 (3) If the message contains any header line starting with "List-" it is
5634 treated as non-personal.
5636 (4) The test for "circular" in the Subject: header line has been removed
5637 because it now seems ill-conceived.
5639 15. Minor typos in src/EDITME comments corrected.
5641 16. Installed latest exipick from John Jetmore.
5643 17. If headers_add on a router specified a text string that was too long for
5644 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5645 of string_sprintf() is now avoided.
5647 18. $message_body_size was not set (it was always zero) when running the DATA
5648 ACL and the local_scan() function.
5650 19. For the "mail" command in an Exim filter, no default was being set for
5651 the once_repeat time, causing a random time value to be used if "once" was
5652 specified. (If the value happened to be <= 0, no repeat happened.) The
5653 default is now 0s, meaning "never repeat". The "vacation" command was OK
5654 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5655 (I found it when inspecting the code).
5657 20. There is now an overall timeout for performing a callout verification. It
5658 defaults to 4 times the callout timeout, which applies to individual SMTP
5659 commands during the callout. The overall timeout applies when there is more
5660 than one host that can be tried. The timeout is checked before trying the
5661 next host. This prevents very long delays if there are a large number of
5662 hosts and all are timing out (e.g. when the network connections are timing
5663 out). The value of the overall timeout can be changed by specifying an
5664 additional sub-option for "callout", called "maxwait". For example:
5666 verify = sender/callout=5s,maxwait=20s
5668 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5669 to the end before writing, but this should make it even safer).
5671 22. Exim was forgetting that it had advertised PIPELINING for the second and
5672 subsequent messages on an SMTP connection. It was also not resetting its
5673 memory on STARTTLS and an internal HELO.
5675 23. When Exim logs an SMTP synchronization error within a session, it now
5676 records whether PIPELINING has been advertised or not.
5678 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5679 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5680 rather than long int.
5682 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5684 26. Added the never_mail option to autoreply.
5690 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5691 crash if the getsockname() call failed; this can happen if a connection is
5692 closed very soon after it is established. The problem was simply in the
5693 order in which certain operations were done, causing Exim to try to write
5694 to the SMTP stream before it had set up the file descriptor. The bug has
5695 been fixed by making things happen in the correct order.
5701 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5702 before the connection was closed, thus losing the rejection response.
5704 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5705 some early Solaris releases, but causes trouble in current releases where
5706 socklen_t is defined.
5708 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5711 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5714 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5715 that releasing the top of it at the end releases what was used for sub-
5716 expansions (unless the block got too big). However, discard this block if
5717 the first thing is a variable or header, so that we can use its block when
5718 it is dynamic (useful for very large $message_headers, for example).
5720 6. Lookups now cache *every* query, not just the most recent. A new, separate
5721 store pool is used for this. It can be recovered when all lookup caches are
5722 flushed. Lookups now release memory at the end of their result strings.
5723 This has involved some general refactoring of the lookup sources.
5725 7. Some code has been added to the store_xxx() functions to reduce the amount
5726 of flapping under certain conditions.
5728 8. log_incoming_interface used to affect only the <= reception log lines. Now
5729 it causes the local interface and port to be added to several more SMTP log
5730 lines, for example "SMTP connection from", and rejection lines.
5732 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5734 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5736 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5737 could be overwritten at the end of the current message (or the start of a
5738 new message if it was set in a HELO ACL). The value is now preserved for
5739 the duration of the SMTP connection.
5741 12. If a transport had a headers_rewrite setting, and a matching header line
5742 contained an unqualified address, that address was qualified, even if it
5743 did not match any rewriting rules. The underlying bug was that the values
5744 of the flags that permit the existence of unqualified sender and recipient
5745 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5746 non-local messages, and by -bnq for local messages) were not being
5747 preserved with the message after it was received.
5749 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5750 "next input=" as part of the text comprising the host identity instead of
5751 the correct text. The code was using the same buffer for two different
5752 strings. However, depending on which order the printing function evaluated
5753 its arguments, the bug did not always show up. Under Linux, for example, my
5754 test suite worked just fine.
5756 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5757 doesn't work with some older versions of Perl. It has been changed to "my",
5758 which in any case is probably the better facility to use.
5760 15. A really picky compiler found some instances of statements for creating
5761 error messages that either had too many or two few arguments for the format
5764 16. The size of the buffer for calls to the DNS resolver has been increased
5765 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5766 for addresses that have a lot of PTR records. This alleviates a problem; it
5767 does not fully solve it.
5769 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5770 buffer now truncates the list and logs the incident, which is the same
5771 action as happens when Exim is looking up a host name and its aliases.
5772 Previously in this situation something unpredictable would happen;
5773 sometimes it was "internal error: store_reset failed".
5775 18. If a server dropped the connection unexpectedly when an Exim client was
5776 using GnuTLS and trying to read a response, the client delivery process
5777 crashed while trying to generate an error log message.
5779 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5780 string, for example:
5782 warn message = H1: something\nH2: something
5784 the text was added as a single header line from Exim's point of view
5785 though it ended up OK in the delivered message. However, searching for the
5786 second and subsequent header lines using $h_h2: did not work. This has been
5787 fixed. Similarly, if a system filter added multiple headers in this way,
5788 the routers could not see them.
5790 20. Expanded the error message when iplsearch is called with an invalid key to
5791 suggest using net-iplsearch in a host list.
5793 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5794 ACLs are no longer actually imposed (and a message to that effect is
5797 22. If a "gecos" field in a passwd entry contained escaped characters, in
5798 particular, if it contained a \" sequence, Exim got it wrong when building
5799 a From: or a Sender: header from that name. A second bug also caused
5800 incorrect handling when an unquoted " was present following a character
5801 that needed quoting.
5803 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5804 was not being matched caselessly.
5806 24. Arranged for all hyphens in the exim.8 source to be escaped with
5809 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5810 properly. Recipient callout cache records were still being keyed to include
5811 the sender, even when use_sender was set false. This led to far more
5812 callouts that were necessary. The sender is no longer included in the key
5813 when use_sender is false.
5815 26. Added "control = submission" modifier to ACLs.
5817 27. Added the ${base62d: operator to decode base 62 numbers.
5819 28. dnsdb lookups can now access SRV records.
5821 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5822 the configuration file.
5824 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5825 (-v) mode. This makes the output for a verbose queue run more intelligible.
5827 31. Added a use_postmaster feature to recipient callouts.
5829 32. Added the $body_zerocount variable, containing the number of binary zero
5830 bytes in the message body.
5832 33. The time of last modification of the "new" subdirectory is now used as the
5833 "mailbox time last read" when there is a quota error for a maildir
5836 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5838 35. Added +ignore_unknown as a special item in host lists.
5840 36. Code for decoding IPv6 addresses in host lists is now included, even if
5841 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5842 address was recognized as an IP address, but was then not correctly decoded
5843 into binary, causing unexpected and incorrect effects when compared with
5850 1. Very minor rewording of debugging text in manualroute to say "list of
5851 hosts" instead of "hostlist".
5853 2. If verify=header_syntax was set, and a header line with an unqualified
5854 address (no domain) and a large number of spaces between the end of the
5855 name and the colon was received, the reception process suffered a buffer
5856 overflow, and (when I tested it) crashed. This was caused by some obsolete
5857 code that should have been removed. The fix is to remove it!
5859 3. When running in the test harness, delay a bit after writing a bounce
5860 message to get a bit more predictability in the log output.
5862 4. Added a call to search_tidyup() just before forking a reception process. In
5863 theory, someone could use a lookup in the expansion of smtp_accept_max_
5864 per_host which, without the tidyup, could leave open a database connection.
5866 5. Added the variables $recipient_data and $sender_data which get set from a
5867 lookup success in an ACL "recipients" or "senders" condition, or a router
5868 "senders" option, similar to $domain_data and $local_part_data.
5870 6. Moved the writing of debug_print from before to after the "senders" test
5873 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5874 problems for message scanning, either using a data ACL, or using
5875 local_scan() because the Received: header was not generated till after they
5876 were called (in order to set the time as the time of reception completion).
5877 I have revised the way this works. The header is now generated after the
5878 body is received, but before the ACL or local_scan() are called. After they
5879 are run, the timestamp in the header is updated.
5885 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5886 before starting a queue runner without re-exec. This happened only when
5887 deliver_drop_privilege was set or when the Exim user was set to root. The
5888 effect of the bug was that timeouts during subsequent deliveries caused
5889 crashes instead of being properly handled. The handler is now left at its
5890 default (and expected) setting.
5892 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5893 message, again when deliver_drop_privilege is set or Exim is run as root.
5894 The bug described in (1) was not present in this case, but the tidying up
5895 of the other signals was missing. I have made the two cases consistent.
5897 3. The ignore_target_hosts setting on a manualroute router was being ignored
5898 for hosts that were looked up using the /MX notation.
5900 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5903 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5904 operated on the sender address. After changing the $sender_address to <>
5905 for the sender address verify, Exim was re-instated it as the original
5906 (before rewriting) address, but remembering that it had rewritten it, so it
5907 wasn't rewriting it again. This bug also had the effect of breaking the
5908 sender address verification caching when the sender address was rewritten.
5910 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5911 This has been changed so that if the ip literal address matches
5912 ignore_target_hosts, the router declines.
5914 7. Added expansion conditions match_domain, match_address, and match_local_
5915 part (NOT match_host).
5917 8. The placeholder for the Received: header didn't have a length field set.
5919 9. Added code to Exim itself and to exim_lock to test for a specific race
5920 condition that could lead to file corruption when using MBX delivery. The
5921 issue is with the lockfile that is created in /tmp. If this file is removed
5922 after a process has opened it but before that process has acquired a lock,
5923 there is the potential for a second process to recreate the file and also
5924 acquire a lock. This could lead to two Exim processes writing to the file
5925 at the same time. The added code performs the same test as UW imapd; it
5926 checks after acquiring the lock that its file descriptor still refers to
5927 the same named file.
5929 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5930 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5933 11. Added the smtp_active_hostname option. If used, this will typically be made
5934 to depend on the incoming interface address. Because $interface_address is
5935 not set up until the daemon has forked a reception process, error responses
5936 that can happen earlier (such as "too many connections") no longer contain
5939 12. If an expansion in a condition on a "warn" statement fails because a lookup
5940 defers, the "warn" statement is abandoned, and the next ACL statement is
5941 processed. Previously this caused the whole ACL to be aborted.
5943 13. Added the iplsearch lookup type.
5945 14. Added ident_timeout as a log selector.
5947 15. Added tls_certificate_verified as a log selector.
5949 16. Added a global option tls_require_ciphers (compare the smtp transport
5950 option of the same name). This controls incoming TLS connections.
5952 17. I finally figured out how to make tls_require_ciphers do a similar thing
5953 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5954 before starting the TLS session.
5956 18. Tabs are now shown as \t in -bP output.
5958 19. If the log selector return_path_on_delivery was set, Exim crashed when
5959 bouncing a message because it had too many Received: header lines.
5961 20. If two routers both had headers_remove settings, and the first one included
5962 a superfluous trailing colon, the final name in the first list and the
5963 first name in the second list were incorrectly joined into one item (with a
5964 colon in the middle).
5970 1. Added -C and -D options to the exinext utility, mainly to make it easier
5971 to include in the automated testing, but these could be helpful when
5972 multiple configurations are in use.
5974 2. The exinext utility was not formatting the output nicely when there was
5975 an alternate port involved in the retry record key, nor when there was a
5976 message id as well (for retries that were specific to a specific message
5977 and a specific host). It was also confused by IPv6 addresses, because of
5978 the additional colons they contain. I have fixed the IPv4 problem, and
5979 patched it up to do a reasonable job for IPv6.
5981 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5982 delivery, the log line now contains "pipelined" if PIPELINING was used.
5984 4. An SMTP transport process used to panic and die if the bind() call to set
5985 an explicit outgoing interface failed. This has been changed; it is now
5986 treated in the same way as a connect() failure.
5988 5. A reference to $sender_host_name in the part of a conditional expansion
5989 that was being skipped was still causing a DNS lookup. This no longer
5992 6. The def: expansion condition was not recognizing references to header lines
5993 that used bh_ and bheader_.
5995 7. Added the _cache feature to named lists.
5997 8. The code for checking quota_filecount in the appendfile transport was
5998 allowing one more file than it should have been.
6000 9. For compatibility with Sendmail, the command line option
6008 and sets the incoming protocol and host name (for trusted callers). The
6009 host name and its colon can be omitted when only the protocol is to be set.
6010 Note the Exim already has two private options, -pd and -ps, that refer to
6011 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6012 "s", but I don't think that's a major issue.
6014 10. A number of refactoring changes to the code, none of which should affect
6017 (a) The number of logging options was getting close to filling up the
6018 32-bit word that was used as a bit map. I have split them into two classes:
6019 those that are passed in the argument to log_write(), and those that are
6020 only ever tested independently outside of that function. These are now in
6021 separate 32-bit words, so there is plenty of room for expansion again.
6022 There is no change in the user interface or the logging behaviour.
6024 (b) When building, for example, log lines, the code previously used a
6025 macro that called string_cat() twice, in order to add two strings. This is
6026 not really sufficiently general. Furthermore, there was one instance where
6027 it was actually wrong because one of the argument was used twice, and in
6028 one call a function was used. (As it happened, calling the function twice
6029 did not affect the overall behaviour.) The macro has been replaced by a
6030 function that can join an arbitrary number of extra strings onto a growing
6033 (c) The code for expansion conditions now uses a table and a binary chop
6034 instead of a serial search (which was left over from when there were very
6035 few conditions). Also, it now recognizes conditions like "pam" even when
6036 the relevant support is not compiled in: a suitably worded error message is
6037 given if an attempt is made to use such a condition.
6039 11. Added ${time_interval:xxxxx}.
6041 12. A bug was causing one of the ddress fields not to be passed back correctly
6042 from remote delivery subprocesses. The field in question was not being
6043 subsequently used, so this caused to problems in practice.
6045 13. Added new log selectors queue_time and deliver_time.
6047 14. Might have fixed a bug in maildirsizefile handling that threw up
6048 "unexpected character" debug warnings, and recalculated the data
6049 unnecessarily. In any case, I expanded the warning message to give more
6052 15. Added the message "Restricted characters in address" to the statements in
6053 the default ACL that block characters like @ and % in local parts.
6055 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6056 Three changes have been made:
6058 (a) There was a serious bug; a negative response to MAIL caused the whole
6059 recipient domain to be cached as invalid, thereby blocking all messages
6060 to all local parts at the same domain, from all senders. This bug has
6061 been fixed. The domain is no longer cached after a negative response to
6062 MAIL if the sender used is not empty.
6064 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6067 (c) A new callout option, "use_sender" has been added for people who want
6068 the modified behaviour.
6074 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6077 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6078 indeed breaks things for older releases.
6080 3. Added additional logging to the case where there is a problem reading data
6081 from a filter that is running in a subprocess using a pipe, in order to
6082 try to track down a specific problem.
6084 4. Testing facility fudge: when running in the test harness and attempting
6085 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6086 getting "No route to host". Convert this to a timeout.
6088 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6091 6. Some OS don't have socklen_t but use size_t instead. This affects the
6092 fifth argument of getsockopt() amongst other things. This is now
6093 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6094 can be set for individual OS. I have set it for SunOS5, OSF1, and
6095 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6096 some earlier ones do not.
6098 7. Change 4.30/15 was not doing the test caselessly.
6100 8. The standard form for an IPv6 address literal was being rejected by address
6101 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6102 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6103 this, as well as the form without the "IPv6" on the front (but only when
6104 address literals are enabled, of course).
6106 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6108 10. Exim crashed if a message with an empty sender address specified by -f
6109 encountered a router with an errors_to setting. This could be provoked only
6110 by a command such as
6114 where an empty string was supplied; "<>" did not hit this bug.
6116 11. Installed PCRE release 4.5.
6118 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6119 remained set. It is now erased.
6121 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6122 times from message ids (which are base 36 rather than the normal 62).
6124 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6125 were being counted as actual protocol errors, and logged if the log
6126 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6127 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6128 and DATA following a set of rejected RCPTs do not count as protocol errors.
6129 In other words, Exim assumes they were pipelined, though this may not
6130 actually be the case. Of course, in all cases the client gets an
6131 appropriate error code.
6133 15. If a lookup fails in an ACL condition, a message about the failure may
6134 be available; it is used if testing the ACL cannot continue, because most
6135 such messages specify what the cause of the deferral is. However, some
6136 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6137 that caused an old message to be retained and used if a later statement
6138 caused a defer, replacing the real cause of the deferral.
6140 16. If an IP address had so many PTR records that the DNS lookup buffer
6141 was not large enough to hold them, Exim could crash while trying to process
6142 the truncated data. It now detects and logs this case.
6144 17. Further to 4.21/58, another change has been made: if (and only if) the
6145 first line of a message (the first header line) ends with CRLF, a bare LF
6146 in a subsequent header line has a space inserted after it, so as not to
6147 terminate the header.
6149 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6150 unnecessarily, used atoi() instead of strtol(), and didn't check the
6151 termination when getting file sizes from file names by regex.
6153 19. Completely re-implemented the support for maildirsize files, in the light
6154 of a number of problems with the previous contributed implementation
6155 (4.30/29). In particular:
6157 . If the quota is zero, the maildirsize file is maintained, but no quota is
6160 . If the maildir directory does not exist, it is created before any attempt
6161 to write a maildirsize file.
6163 . The quota value in the file is just a cache; if the quota is changed in
6164 the transport, the new value overrides.
6166 . A regular expression is available for excluding directories from the
6169 20. The autoreply transport checks the characters in options that define the
6170 message's headers; it allows continued headers, but it was checking with
6171 isspace() after an embedded newline instead of explicitly looking for a
6174 21. If all the "regular" hosts to which an address was routed had passed their
6175 expiry times, and had not reached their retry times, the address was
6176 bounced, even if fallback hosts were defined. Now Exim should go on to try
6179 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6180 equivalent code in the SMTP transport. Some hosts send humungous responses
6181 to HELO/EHLO, more than 1024 it seems.
6183 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6184 gives compiler warnings in some environments. I've now done it "properly",
6187 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6188 (because of problems with the built-in one) was declared to return uschar *
6189 instead of char *, causing compiler failure.
6191 25. Fixed a file descriptor leak when processing alias/forward files.
6193 26. Fixed a minor format string issue in dbfn.c.
6195 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6197 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6198 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6201 29. When a sender address is verified, it is cached, to save repeating the test
6202 when there is more than one recipient in a message. However, when the
6203 verification involves a callout, it is possible for different callout
6204 options to be set for different recipients. It is too complicated to keep
6205 track of this in the cache, so now Exim always runs a verification when a
6206 callout is required, relying on the callout cache for the optimization.
6207 The overhead is duplication of the address routing, but this should not be
6210 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6211 to be verified with callout=postmaster, and the main callout worked but the
6212 postmaster check failed, the verification correctly failed. However, if a
6213 subsequent RCPT command asked for sender verification *without* the
6214 postmaster check, incorrect caching caused this verification also to fail,
6217 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6218 it was not caching the DNS options (qualify_single, search_parents) that
6219 were used when the lookup failed. A subsequent lookup with different
6220 options therefore always gave the same answer, though there were cases
6221 where it should not have. (Example: a "domains = !$mx_any" option on a
6222 dnslookup router: the "domains" option is always processed without any
6223 widening, but the router might have qualify_single set.) Now Exim uses the
6224 cached value only when the same options are set.
6226 32. Added John Jetmore's "exipick" utility to the distribution.
6228 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6229 than a timeout (e.g. a certificate is required, and is not provided), an
6230 Exim server now closes the connection immediately. Previously it waited for
6231 the client to close - but if the client is SSL, it seems that they each
6232 wait for each other, leading to a delay before one of them times out.
6234 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6235 maintained 0.8.x compatibility because I don't think many are using it, and
6236 it is clearly obsolete.
6238 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6241 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6242 client certificate was expired. A simple patch fixes this, though I don't
6243 understand the full logic of why the verify callback is called multiple
6246 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6247 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6248 which causes problems with some clients (such as the Certicom SSL Plus
6249 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6250 disables the coutermeasure allowing Eudora to connect."
6252 38. Exim was not checking that a write() to a log file succeeded. This could
6253 lead to Bad Things if a log got too big, in particular if it hit a file
6254 size limit. Exim now panics and dies if it cannot write to a log file, just
6255 as it does if it cannot open a log file.
6257 39. Modified OS/Makefile-Linux so that it now contains
6259 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6261 The two -D definitions ensure that Exim is compiled with large file
6262 support, which makes it possible to handle log files that are bigger than
6265 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6266 instance) a domain was checked against a named list that involved a lookup,
6267 causing $domain_data to be set, then another domain was checked against the
6268 same list, then the first domain was re-checked, the value of $domain_data
6269 after the final check could be wrong. In particular, if the second check
6270 failed, it could be set empty. This bug probably also applied to
6273 41. The strip_trailing_dot option was not being applied to the address given
6274 with the -f command-line option.
6276 42. The code for reading a message's header from the spool was incrementing
6277 $received_count, but never initializing it. This meant that the value was
6278 incorrect (doubled) while delivering a message in the same process in which
6279 it was received. In the most common configuration of Exim, this never
6280 happens - a fresh exec is done - but it can happen when
6281 deliver_drop_privilege is set.
6283 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6284 it now includes up to 150 characters of the unexpected data in the log
6287 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6288 and building data strings. The size of both of these buffers was 10 000
6289 bytes - far larger than anybody would *ever* want, thought I. Needless to
6290 say, somebody hit the limit. I have increased the maximum line length to
6291 20 000 and the maximum data length of concatenated lines to 100 000. I have
6292 also fixed two bugs, because there was no checking on these buffers. Tsk,
6293 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6294 buffer is too small.
6296 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6297 lsearch lookups. Now it does.
6299 46. When parsing a route_list item in a manualroute router, a fixed-length
6300 buffer was used for the list of hosts. I made this 1024 bytes long,
6301 thinking that nobody would ever have a list of hosts that long. Wrong.
6302 Somebody had a whole pile of complicated expansion conditions, and the
6303 string was silently truncated, leading to an expansion error. It turns out
6304 that it is easier to change to an unlimited length (owing to other changes
6305 that have happened since this code was originally written) than to build
6306 structure for giving a limitation error. The length of the item that
6307 expands into the list of hosts is now unlimited.
6309 47. The lsearch lookup could not handle data where the length of text line was
6310 more than 4095 characters. Such lines were truncated, leading to shortened
6311 data being returned. It should now handle lines of any length.
6313 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6314 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6317 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6318 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6319 no longer happens in comment lines. A list of replacements is now placed
6320 at the head of all of the source files, except those whose only change is
6321 to replace PERL_COMMAND in the very first #! line.
6323 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6324 messages on the queue, with a bottom-up merge sort, using code contributed
6325 by Michael Haardt. This should make operations like -bp somewhat faster on
6326 large queues. It won't affect queue runners, except when queue_run_in_order
6329 51. Installed eximstats 1.31 in the distribution.
6331 52. Added support for SRV lookups to the dnslookup router.
6333 53. If an ACL referred to $message_body or $message_body_end, the value was not
6334 reset for any messages that followed in the same SMTP session.
6336 54. The store-handling optimization for building very long strings was not
6337 differentiating between the different store pools. I don't think this
6338 actually made any difference in practice, but I've tidied it.
6340 55. While running the routers to verify a sender address, $sender_address
6341 was still set to the sender address. This is wrong, because when routing to
6342 send a bounce to the sender, it would be empty. Therefore, I have changed
6343 it so that, while verifying a sender address, $sender_address is set to <>.
6344 (There is no change to what happens when verifying a recipient address.)
6346 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6347 target A or AAAA records (if not already returned) without resetting the
6348 qualify_single or search_parents options of the DNS resolver. These are
6349 inappropriate in this case because the targets of MX and SRV records must
6350 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6351 target that, when qualified, matched something in the local domain. These
6352 two options are now turned off when doing these lookups.
6354 57. It seems that at least some releases of Reiserfs (which does not have the
6355 concept of a fixed number of inodes) returns zero and not -1 for the
6356 number of available inodes. This interacted badly with check_spool_inodes,
6357 which assumed that -1 was the "no such thing" setting. What I have done is
6358 to check that the total number of inodes is greater than zero before doing
6359 the test of how many are available.
6361 58. When a "warn" ACL statement has a log_message modifier, the message is
6362 remembered, and not repeated. This is to avoid a lot of repetition when a
6363 message has many recipients that cause the same warning to be written.
6364 However, Exim was preserving the list of already written lines for an
6365 entire SMTP session, which doesn't seem right. The memory is now reset if a
6366 new message is started.
6368 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6369 debugging output unless log_rewrite was also set.
6371 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6372 of (int)(handle) when we know that handle contains (void *)(-1).
6374 61. The Exim daemon panic-logs an error return when it closes the incoming
6375 connection. However "connection reset by peer" seems to be common, and
6376 isn't really an error worthy of noting specially, so that particular error
6379 62. When Exim is trying to find all the local interfaces, it used to panic and
6380 die if the ioctl to get the interface flags failed. However, it seems that
6381 on at least one OS (Solaris 9) it is possible to have an interface that is
6382 included in the list of interfaces, but for which you get a failure error
6383 for this call. This happens when the interface is not "plumbed" into a
6384 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6385 failure of the "get flags" call assumes that the interface is down.
6387 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6388 makes life easier for people who are doing arithmetic on fields extracted
6389 from dates, where you often get leading zeros that should not be
6390 interpreted as octal.
6392 64. Added qualify_domain to the redirect router, to override the global
6395 65. If a pathologically long header line contained very many addresses (the
6396 report of this problem mentioned 10 000) and each of them was rewritten,
6397 Exim could use up a very large amount of memory. (It kept on making new
6398 copies of the header line as it rewrote, and never released the old ones.)
6399 At the expense of a bit more processing, the header rewriting function has
6400 been changed so that it no longer eats memory in this way.
6402 66. The generation of the Received: header has been moved from the time that a
6403 message starts to be received, to the time that it finishes. The timestamp
6404 in the Received: header should now be very close to that of the <= log
6405 line. There are two side-effects of this change:
6407 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6408 logged header lines no longer include the local Received: line, because
6409 it has not yet been created. The same applies to a copy of the message
6410 that is returned to a non-SMTP sender when a message is rejected.
6412 (b) When a filter file is tested using -bf, no additional Received: header
6413 is added to the test message. After some thought, I decided that this
6416 This change does not affect the value of $received_for. It is still set
6417 after address rewriting, but before local_scan() is called.
6419 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6421 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6422 gave an unhelpful panic error message, and a defer error. I have managed to
6423 change this behaviour so that it now rejects any supplied certificate,
6424 which seems right, as the list of acceptable certificates is empty.
6426 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6427 gave an unhelpful defer error. I have not managed to make this reject any
6428 supplied certificates, but the error message it gives is "no certificate
6429 supplied", which is not helpful.
6431 70. exigrep's output now also includes lines that are not associated with any
6432 message, but which match the given pattern. Implemented by a patch from
6433 Martin Sluka, which also tidied up the Perl a bit.
6435 71. Recipient callout verification, like sender verification, was using <> in
6436 the MAIL FROM command. This isn't really the right thing, since the actual
6437 sender may affect whether the remote host accepts the recipient or not. I
6438 have changed it to use the actual sender in the callout; this means that
6439 the cache record is now keyed on a recipient/sender pair, not just the
6440 recipient address. There doesn't seem to be a real danger of callout loops,
6441 since a callout by the remote host to check the sender would use <>.
6442 [SEE ABOVE: changed after hitting problems.]
6444 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6445 temporary errors. However, in the case of such a code being given after
6446 the end of a data transmission (i.e. after ".") Exim was failing to write
6447 a retry record for the message. (Yes, there was some broken host that was
6448 actually sending 8xx at this point.)
6450 73. An unknown lookup type in a host list could cause Exim to panic-die when
6451 the list was checked. (An example that provoked this was putting <; in the
6452 middle of a list instead of at the start.) If this happened during a DATA
6453 ACL check, a -D file could be left lying around. This kind of configuration
6454 error no longer causes Exim to die; instead it causes a defer error. The
6455 incident is still logged to the main and panic logs.
6457 74. Buglet left over from Exim 3 conversion. The message "too many messages
6458 in one connection" was written to the rejectlog but not the mainlog, except
6459 when address rewriting (yes!) was being logged.
6461 75. Added write_rejectlog option.
6463 76. When a system filter was run not as root (that is, when system_filter_user
6464 was set), the values of the $n variables were not being returned to the
6465 main process; thus, they were not subsequently available in the $sn
6468 77. Added +return_path_on_delivery log selector.
6470 78. A connection timeout was being treated differently from recipients deferred
6471 when testing hosts_max_try with a message that was older than the host's
6472 retry timeout. (The host should not be counted, thus allowing all hosts to
6473 be tried at least once before bouncing.) This may have been the cause of an
6474 occasionally reported bug whereby a message would remain on the queue
6475 longer than the retry timeout, but would be bounced if a delivery was
6476 forced. I say "may" because I never totally pinned down the problem;
6477 setting up timeout/retry tests is difficult. See also the next item.
6479 79. The ultimate address timeout was not being applied to errors that involved
6480 a combination of host plus message (for example, a timeout on a MAIL
6481 command). When an address resolved to a number of possible hosts, and they
6482 were not all tried for each delivery (e.g. because of hosts_max_try), a
6483 message could remain on the queue longer than the retry timeout.
6485 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6488 81. Fixed an obscure SMTP outgoing bug which required at least the following
6489 conditions: (a) there was another message waiting for the same server;
6490 (b) the server returned 5xx to all RCPT commands in the first message so
6491 that the message was not completed; (c) the server dropped the connection
6492 or gave a negative response to the RSET that Exim sends to abort the
6493 transaction. The observed case was a dropped connection after DATA that had
6494 been sent in pipelining mode. That is, the server had advertised PIPELINING
6495 but was not implementing it correctly. The effect of the bug was incorrect
6496 behaviour, such as trying another host, and this could lead to a crash.
6502 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6503 and daemon.c were passed as pointers to ints; they should have been
6504 pointers to socklen_t variables (which are typically unsigned ints).
6506 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6509 3. Fixed a really odd bug that affected only the testing scheme; patching a
6510 certain fixed string in the binary changed the value of another string that
6511 happened to be identical to the end of the original first string.
6513 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6514 name", it returns that address as the IP address. On some operating
6515 systems (e.g. Solaris), it also passes back the IP address string as the
6516 "host name". However, on others (e.g. Linux), it passes back an empty
6517 string. Exim wasn't checking for this, and was changing the host name to an
6518 empty string, assuming it had been canonicalized.
6520 5. Although rare, it is permitted to have more than one PTR record for a given
6521 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6522 all the names associated with an address, because they do in Solaris.
6523 However, it seems that they do not in Linux for data that comes from the
6524 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6525 I found this out when I moved to a new Linux workstation and tried to run
6526 the Exim test suite.
6528 To get round this problem I have changed the code so that it now does its
6529 own call to the DNS to look up PTR records when searching for a host name.
6530 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6531 addresses that are only in /etc/hosts are still found.
6533 This behaviour is, however, controlled by an option called host_lookup_
6534 order, which defaults to "bydns:byaddr". If people want to use the other
6535 order, or indeed, just use one or the other means of lookup, they can
6536 specify it in this variable.
6538 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6539 some operating systems, this comes back from gethostbyaddr() as an empty
6540 string, and this is what Exim used to test for. However, it seems that in
6541 other systems, "." is yielded. Exim now tests for this case too.
6543 7. The values of check_spool_space and check_log_space are now held internally
6544 as a number of kilobytes instead of an absolute number of bytes. If a
6545 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6546 kilobyte. This means that much larger values can be stored.
6548 8. Exim monitor: an attempt to get the action menu when not actually pointing
6549 at a message produces an empty menu entitled "No message selected". This
6550 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6551 no entries in it ("Shell widget menu has zero width and/or height"). So I
6552 have added a single, blank menu entry in this case.
6554 9. Added ${quote_local_part.
6556 10. MIME decoding is now applied to the contents of Subject: header lines when
6559 11. Now that a reference to $sender_host_address automatically causes a reverse
6560 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6561 host lookup before query-style lookups in lists that might use this
6562 variable. This has therefore been abolished, and the "net-" prefix is no
6563 longer necessary for query-style lookups.
6565 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6566 have been a typo for LFLAGS, so it has been changed.
6568 13. The install script calls Exim with "-C /dev/null" in order to find the
6569 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6570 to be output. However, since Exim outputs its version number before the
6571 error, it didn't break the script. It just looked ugly. I fixed this by
6572 always allowing "-C /dev/null" if the caller is root.
6574 14. Ignore overlarge ACL variable number when reading spool file - insurance
6575 against a later release with more variables having written the file.
6577 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6578 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6579 this, as well as the form without the "IPv6" on the front.
6581 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6582 OS/Makefile-Darwin file.
6584 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6585 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6587 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6588 variable. However, some OS use ints for inodes. I've added cast to long int
6589 to get rid of the compiler warning.
6591 19. I had forgotten to lock out "/../" in configuration file names when
6592 ALT_CONFIG_PREFIX was set.
6594 20. Routers used for verification do not need to specify transports. However,
6595 if such a router generated a host list, and callout was configured, Exim
6596 crashed, because it could not find a port number from the (non-existent)
6597 transport. It now assumes port 25 in this circumstance.
6599 21. Added the -t option to exigrep.
6601 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6602 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6603 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6604 them is set, LOOKUP_LSEARCH is forced.
6606 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6608 24. Added sender and host information to the "rejected by local_scan()" log
6609 line; previously there was no indication of these.
6611 25. Added .include_if_exists.
6613 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6614 Linux. It turns out that not all file systems support this. Apparently some
6615 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6616 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6617 not supported on the file descriptor, is now ignored when Exim is trying to
6618 sync a directory. This applies only to Linux.
6620 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6622 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6623 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6626 29. Added support for maildirsize files from supplied patch (modified a bit).
6628 30. The use of :fail: followed by an empty string could lead Exim to respond to
6629 sender verification failures with (e.g.):
6631 550 Verification failed for <xxx>
6632 550 Sender verify failed
6634 where the first response line was missing the '-' that indicates it is not
6635 the final line of the response.
6637 31. The loop for finding the name of the user that called Exim had a hardwired
6638 limit of 10; it now uses the value of finduser_retries, which is used for
6639 all other user lookups.
6641 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6644 33. Exim was neglecting to zero errno before one call of strtol() when
6645 expanding a string and expecting an integer value. On some systems this
6646 resulted in spurious "integer overflow" errors. Also, it was casting the
6647 result into an int without checking.
6649 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6650 did not work. The code looks as if it has *never* worked, though it appears
6651 to have been documented since at least release 1.62. I have made it work.
6653 35. The "timeout_DNS" error in retry rules, also documented since at least
6654 1.62, also never worked. As it isn't clear exactly what this means, and
6655 clearly it isn't a major issue, I have abolished the feature by treating it
6656 as "timeout", and writing a warning to the main and panic logs.
6658 36. The display of retry rules for -brt wasn't always showing the error code
6661 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6662 timeout_connect_A, timeout_connect_MX.
6664 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6665 to the empty sender.
6667 39. The daemon was not analysing the content of -oX till after it had closed
6668 stderr and disconnected from the controlling terminal. This meant that any
6669 syntax errors were only noted on the panic log, and the return code from
6670 the command was 0. By re-arranging the code a little, I've made the
6671 decoding happen first, so such errors now appear on stderr, and the return
6672 code is 1. However, the actual setting up of the sockets still happens in
6673 the disconnected process, so errors there are still only recorded on the
6676 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6677 connections (as happens on some IP stacks) was logged at start up time as
6678 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6679 it from "IPv6 and IPv4", which means that two separate sockets are being
6682 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6683 says whether AF_INET or AF_INET6 was passed as an argument.
6685 42. Exiwhat output was messed up when time zones were included in log
6688 43. Exiwhat now gives more information about the daemon's listening ports,
6689 and whether -tls-on-connect was used.
6691 44. The "port" option of the smtp transport is now expanded.
6693 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6694 silently ignored. Now an error message is written to the main and panic
6697 46. There's a new ACL modifier called "logwrite" which writes to a log file
6698 as soon as it is encountered.
6700 47. Added $local_user_uid and $local_user_gid at routing time.
6702 48. Exim crashed when trying to verify a sender address that was being
6705 49. Exim was recognizing only a space character after ".include". It now also
6706 recognizes a tab character.
6708 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6709 extracting the relevant information from the specification. The man page no
6710 longer contains scrambled data for the -d option, and I've added a section
6711 at the front about calling Exim under different names.
6713 51. Added "extra_headers" argument to the "mail" command in filter files.
6715 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6718 53. Installed eximstats 1.29.
6720 54. Added transport_filter_timeout as a generic transport option.
6722 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6723 Cc: header lines. This was required by RFC 822, but it not required by RFC
6726 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6727 incoming messages that did not have them. Now it does so only if the
6728 message originates locally, that is, if there is no associated remote host
6729 address. When Resent- header lines are present, this applies to the Resent-
6730 lines rather than the non-Resent- lines.
6732 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6733 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6735 58. Messages for configuration errors now include the name of the main
6736 configuration file - useful now that there may be more than one file in a
6737 list (.included file names were always shown).
6739 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6740 for those rare installations that do not start the daemon as root or run it
6741 setuid root. I've cut out the call to initgroups() if the daemon is not
6744 60. The Exim user and group can now be bound into the binary as text strings
6745 that are looked up at the start of Exim's processing.
6747 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6749 62. Added $mailstore_basename variable.
6751 63. Installed patch to sieve.c from Michael Haardt.
6753 64. When Exim failed to open the panic log after failing to open the main log,
6754 the original message it was trying to log was written to stderr and debug
6755 output, but if they were not available (the usual case in production), it
6756 was lost. Now it is written to syslog before the two lines that record the
6757 failures to open the logs.
6759 65. Users' Exim filters run in subprocesses under the user's uid. It is
6760 possible for a "deliver" command or an alias in a "personal" command to
6761 provoke an address rewrite. If logging of address rewriting is configured,
6762 this fails because the process is not running as root or exim. There may be
6763 a better way of dealing with this, but for the moment (because 4.30 needs
6764 to be released), I have disabled address rewrite logging when running a
6765 filter in a non-root, non-exim process.
6771 1. The buildconfig auxiliary program wasn't quoting the value set for
6772 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6773 not defined. This bug was masked in 4.22 by the effect that was fixed in
6776 2. Some messages that were rejected after a message id was allocated were
6777 shown as "incomplete" by exigrep. It no longer does this for messages that
6778 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6780 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6781 have allow_domain_literals set, the ID did not get logged in the <= line.
6782 Domain literals are now always recognized in Message-ID: header lines.
6784 4. The first argument for a ${extract expansion item is the key name or field
6785 number. Leading and trailing spaces in this item were not being ignored,
6786 causing some misleading effects.
6788 5. When deliver_drop_privilege was set, single queue runner processes started
6789 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6790 same command in the process it spins off) were not dropping privilege.
6792 6. When the daemon running as "exim" started a queue runner, it always
6793 re-executed Exim in the spun-off process. This is a waste of effort when
6794 deliver_drop_privilege is set. The new process now just calls the
6795 queue-runner function directly.
6801 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6804 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6805 ignored. Though the use of -oP was forcing the writing of a pid file, it
6806 was always written to the default place.
6808 3. If the message "no IP address found for host xxxx" is generated during
6809 incoming verification, it is now followed by identification of the incoming
6810 connection (so you can more easily find what provoked it).
6812 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6814 5. Added some features to "harden" Exim a bit more against certain attacks:
6816 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6817 be put in Local/Makefile. This is like the never_users runtime option,
6818 but it cannot be overridden. The default setting is "root".
6820 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6821 prefix string with which any file named in a -C command line option
6824 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6825 is retained for -C and -D only if the caller of Exim is root. Without
6826 it, the exim user may also use -C and -D and retain privilege.
6828 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6829 command line option is disabled.
6831 6. Macro names set by the -D option must start with an upper case letter, just
6832 like macro names defined in the configuration file.
6834 7. Added "dereference=" facility to LDAP.
6836 8. Two instances of the typo "uknown" in the source files are fixed.
6838 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6839 the Configure-Makefile script screwed up while processing it.
6841 10. Incorporated PCRE 4.4.
6843 11. The SMTP synchronization check was not operating right at the start of an
6844 SMTP session. For example, it could not catch a HELO sent before the client
6845 waited for the greeting. There is now a check for outstanding input at the
6846 point when the greeting is written. Because of the duplex, asynchronous
6847 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6848 way, but not yet received, when the check is performed.
6850 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6851 on TCP/IP sockets, because this apparently causes some broken clients to
6854 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6855 unchanged) from the Cygwin maintainer.
6857 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6858 was supported instead of testing for mbx. Effectively a typo.
6860 15. The spa authenticator server code was not checking that the input it
6861 received was valid base64.
6863 16. The debug output line for the "set" modifier in ACLs was not showing the
6864 name of the variable that was being set.
6866 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6868 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6869 lookup on its own. Something else was needed to trigger the lookup. For
6870 example, a match in host_lookup or the need for a host name in a host list.
6871 Now, if $sender_host_name is referenced and the host name has not yet been
6872 looked up, a lookup is performed. If the lookup fails, the variable remains
6873 empty, and $host_lookup_failed is set to "1".
6875 19. Added "eqi" as a case-independent comparison operator.
6877 20. The saslauthd authentication condition could segfault if neither service
6878 nor realm was specified.
6880 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6881 error message that was logged was misleading, and incoming SMTP
6882 connections were dropped. The message is now more accurate, and temporary
6883 errors are given to SMTP connections.
6885 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6886 (except RSET) with a 421 error, until QUIT is received. However, it was
6887 failing to send a response to QUIT.
6889 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6890 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6891 ACL is now run after the helo_try_verify_hosts code.
6893 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6894 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6895 case-independent, so other case variants are also recognized). Apparently
6896 some systems use these upper case variants.
6898 25. If more than two messages were waiting for the same host, and a transport
6899 filter was specified for the transport, Exim sent two messages over the
6900 same TCP/IP connection, and then failed with "socket operation on non-
6901 socket" when it tried to send the third.
6903 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6905 27. The extern definition of crypt16() in expand.c was not being excluded when
6906 the OS had its own crypt16() function.
6908 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6909 as a preferred synonym for return_size_limit, both as an option and as an
6912 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6914 30. Changed the default configuration ACL to relax the local part checking rule
6915 for addresses that are not in any local domains. For these addresses,
6916 slashes and pipe symbols are allowed within local parts, but the sequence
6917 /../ is explicitly forbidden.
6919 31. SPA server authentication was not clearing the challenge buffer before
6922 32. log_message in a "warn" ACL statement was writing to the reject log as
6923 well as to the main log, which contradicts the documentation and doesn't
6924 seem right (because no rejection is happening). So I have stopped it.
6926 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6927 However, I am unable to do any testing of this.
6929 34. Fixed an infelicity in the appendfile transport. When checking directories
6930 for a mailbox, to see if any needed to be created, it was accidentally
6931 using path names with one or more superfluous leading slashes; tracing
6932 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6934 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6935 contents are added to the log line that is written for every discarded
6936 recipient. (Previously a log_message setting was ignored.)
6938 36. The ${quote: operator now quotes the string if it is empty.
6940 37. The install script runs exim in order to find its version number. If for
6941 some reason other than non-existence or emptiness, which it checks, it
6942 could not run './exim', it was installing it with an empty version number,
6943 i.e. as "exim-". This error state is now caught, and the installation is
6946 38. An argument was missing from the function that creates an error message
6947 when Exim fails to connect to the socket for saslauthd authentication.
6948 This could cause Exim to crash, or give a corrupted message.
6950 39. Added isip, isip4, and isip6 to ${if conditions.
6952 40. The ACL variables $acl_xx are now saved with the message, and can be
6953 accessed later in routers, transports, and filters.
6955 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6956 strings in the file are not string-expanded.
6958 42. If a MAIL command specified a SIZE value that was too large to fit into an
6959 int variable, the check against message_size_limit failed. Such values are
6960 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6961 day this will have to be increased, but I don't think I want to be around
6962 when emails are that large.
6969 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6970 iconv() is not standard in FreeBSD.
6972 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6973 IPv6 enabled. The observed symptom was a segmentation fault on return from
6974 the function os_common_find_running_interfaces() in src/os.c.
6976 3. In the check_special_case() function in daemon.c I had used "errno" as an
6977 argument name, which causes warnings on some systems. This was basically a
6978 typo, since it was named "eno" in the comments!
6980 4. The code that waits for the clock to tick (at a resolution of some fraction
6981 of a second) so as to ensure message-id uniqueness was always waiting for
6982 at least one whole tick, when it could have waited for less. [This is
6983 almost certainly not relevant at current processor speeds, where it is
6984 unlikely to ever wait at all. But we try to future-proof.]
6986 5. The function that sleeps for a time interval that includes fractions of a
6987 second contained a race. It did not block SIGALRM between setting the
6988 timer, and suspending (a couple of lines later). If the interval was short
6989 and the sigsuspend() was delayed until after it had expired, the suspension
6990 never ended. On busy systems this could lead to processes getting stuck for
6993 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6994 process, before it forks any delivery processes. The open lookup caching
6995 mechanism meant that the open file or database connection was passed into
6996 the delivery process. The problem was that delivery processes always tidy
6997 up cached lookup data. This could cause a problem for the next delivery
6998 process started by the queue runner, because the external queue runner
6999 process does not know about the closure. So the next delivery process
7000 still has data in the lookup cache. In the case of a file lookup, there was
7001 no problem because closing a file descriptor in a subprocess doesn't affect
7002 the parent. However, if the lookup was caching a connection to a database,
7003 the connection was closed, and the second delivery process was likely to
7004 see errors such as "PGSQL: query failed: server closed the connection
7005 unexpectedly". The problem has been fixed by closing all cached lookups
7006 in a queue runner before running a delivery process.
7008 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7009 seem to have the "const" qualifier which it has on other OS. I've
7012 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7013 *of the same type* (client or server) with the same public name that an
7014 error should be diagnosed.
7016 9. When Exim looked up a host name for an IP address, but failed to find the
7017 original IP address when looking up the host name (a safety check), it
7018 output the message "<ip address> does not match any IP for NULL", which was
7019 confusing, to say the least. The bug was that the host name should have
7020 appeared instead of "NULL".
7022 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7023 user that is built into the binary, and the -C or -D options is used, root
7024 privilege is dropped before the configuration file is read. In addition,
7025 logging is switched to stderr instead of the normal log files. If the
7026 configuration then re-defines the Exim user, the unprivileged environment
7027 is probably not what is expected, so Exim logs a panic warning message (but
7030 However, if deliver_drop_privilege is set, the unprivileged state may well
7031 be exactly what is intended, so the warning has been cut out in that case,
7032 and Exim is allowed to try to write to its normal log files.
7038 1. smtp_return_error_details was not giving details for temporary sender
7039 or receiver verification errors.
7041 2. Diagnose a configuration error if two authenticators have the same public
7044 3. Exim used not to create the message log file for a message until the first
7045 delivery attempt. This could be confusing when incoming messages were held
7046 for policy or load reasons. The message log file is now created at the time
7047 the message is received, and an initial "Received" line is written to it.
7049 4. The automatically generated man page for command line options had a minor
7050 bug that caused no ill effects; however, a more serious problem was that
7051 the procedure for building the man page automatically didn't always
7052 operate. Consequently, release 4.20 contains an out-of-date version. This
7053 shouldn't happen again.
7055 5. When building Exim with embedded Perl support, the script that builds the
7056 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7057 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7059 6. The freeze_tell option was not being used for messages that were frozen on
7060 arrival, either by an ACL or by local_scan().
7062 7. Added the smtp_incomplete_transaction log selector.
7064 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7065 was accepting AUTH without a new EHLO.
7067 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7068 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7071 10. Logging of TCP/IP connections (when configured) now happens in the main
7072 daemon process instead of the child process, so that the TCP/IP connection
7073 count is more accurate (but it can never be perfect).
7075 11. The use of "drop" in a nested ACL was not being handled correctly in the
7076 outer ACL. Now, if condition failure induced by the nested "drop" causes
7077 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7078 or "require"), the connection is dropped.
7080 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7081 that yield "discard" can now be used with an "accept" or a "discard" verb,
7082 but an error is generated for any others (because I can't see a useful way
7083 to define what should happen).
7085 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7086 are now processed in the same way as lines in the Exim configuration file.
7087 In particular, continuation lines are supported.
7089 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7091 15. Added -ti meaning -t -i.
7093 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7094 domains, and warn by logging if others are found.
7096 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7097 returned from GIFCONF when Exim is trying to find the list of interfaces on
7098 a host. The code in os.c has been modified to copy each ifreq to an aligned
7099 structure in all cases.
7101 Also, in some cases, the returned ifreq's were being copied to a 'struct
7102 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7103 means the last couple of bytes of an IPv6 address could be chopped if the
7104 ifreq contained only a normal sockaddr (14 bytes storage).
7106 18. Named domain lists were not supported in the hosts_treat_as_local option.
7107 An entry such as +xxxx was not recognized, and was treated as a literal
7110 19. Ensure that header lines added by a DATA ACL are included in the reject log
7111 if the ACL subsequently rejects the message.
7113 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7114 MD5 (which is deprecated).
7116 21. When testing a filter file using -bf, Exim was writing a message when it
7117 took the sender from a "From " line in the message, but it was not doing so
7118 when it took $return_path from a Return-Path: header line. It now does.
7120 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7121 with a valid header line field name (a series of printing characters
7122 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7124 23. Changed "disc" in the source to "disk" to conform to the documentation and
7125 the book and for uniformity.
7127 24. Ignore Sendmail's -Ooption=value command line item.
7129 25. When execve() failed while trying to run a command in a pipe transport,
7130 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7131 could be confused with a return value of 69 from the command itself. This
7132 has been changed to 127, the value the shell returns if it is asked to run
7133 a non-existent command. The wording for the related log line suggests a
7134 non-existent command as the problem.
7136 26. If received_header_text expands to an empty string, do not add a Received:
7137 header line to the message. (Well, it adds a token one on the spool, but
7138 marks it "old" so that it doesn't get used or transmitted.)
7140 27. Installed eximstats 1.28 (addition of -nt option).
7142 28. There was no check for failure on the call to getsockname() in the daemon
7143 code. This can fail if there is a shortage of resources on the system, with
7144 ENOMEM, for example. A temporary error is now given on failure.
7146 29. Contrary to the C standard, it seems that in some environments, the
7147 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7148 program. Exim now does this explicitly; it affects the formatting of
7149 timestamps using strftime().
7151 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7152 complaints. I've now initialized all the variables, to avoid this.
7154 32. Header lines added by a system filter were not being "seen" during
7155 transport-time rewrites.
7157 33. The info_callback() function passed to OpenSSL is set up with type void
7158 (*)(SSL *, int, int), as described somewhere. However, when calling the
7159 function (actually a macro) that sets it up, the type void(*)() is
7160 expected. I've put in a cast to prevent warnings from picky compilers.
7162 34. If a DNS black list lookup found a CNAME record, but there were no A
7163 records associated with the domain it pointed at, Exim crashed.
7165 35. If a DNS black list lookup returned more than one A record, Exim ignored
7166 all but the first. It now scans all returned addresses if a particular IP
7167 value is being sought. In this situation, the contents of the
7168 $dnslist_value variable are a list of all the addresses, separated by a
7171 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7172 to accept a host name and all its aliases if the forward lookup for any of
7173 them yielded the IP address of the incoming connection. Now it accepts only
7174 those names whose forward lookup yields the correct IP address. Any other
7175 names are discarded. This closes a loophole whereby a rogue DNS
7176 administrator could create reverse DNS records to break through a
7177 wildcarded host restriction in an ACL.
7179 37. If a user filter or a system filter that ran in a subprocess used any of
7180 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7181 the wrong values were passed to the pipe command ($thisaddress had the
7182 value of $0, $0 had the value of $1, etc). This bug was introduced by
7183 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7185 38. Improved the line breaking for long SMTP error messages from ACLs.
7186 Previously, if there was no break point between 40 and 75 characters, Exim
7187 left the rest of the message alone. Two changes have been made: (a) I've
7188 reduced the minimum length to 35 characters; (b) if it can't find a break
7189 point between 35 and 75 characters, it looks ahead and uses the first one
7190 that it finds. This may give the occasional overlong line, but at least the
7191 remaining text gets split now.
7193 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7194 file descriptors might be low, and that setting 1000 would always raise it.
7195 It turns out that in some environments, the limit is already over 1000 and
7196 that lowering it causes trouble. So now Exim takes care not to decrease it.
7198 40. When delivering a message, the value of $return_path is set to $sender_
7199 address at the start of routing (routers may change the value). By an
7200 oversight, this default was not being set up when an address was tested by
7201 -bt or -bv, which affected the outcome if any router or filter referred to
7204 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7205 log only when "message" or "log_message" are set. However, if one of the
7206 conditions was an address verification, or a call to a nested ACL, the
7207 messages generated by the underlying test were being passed through. This
7208 no longer happens. The underlying message is available in $acl_verify_
7209 message for both "message" and "log_message" expansions, so it can be
7210 passed through if needed.
7212 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7213 new expansion $bh_ to give the encoded byte string without charset
7214 translation. Translation happens only if iconv() is available; HAVE_ICONV
7215 indicates this at build time. HEADERS_CHARSET gives the charset to
7216 translate to; headers_charset can change it in the configuration, and
7217 "headers charset" can change it in an individual filter file.
7219 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7220 that creates RFC 2047 encoded "words" labels them as that charset instead
7221 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7222 expansion operator; (ii) when Exim creates a From: line for a local
7223 message; (iii) when a header line is rewritten to include a "phrase" part.
7225 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7226 buggy, causing it to skip the first lines of messages whose message ID
7227 ended in 'D'. This would not have bitten before Exim release 4.14, because
7228 message IDs were unlikely to end in 'D' before then. The effect was to have
7229 incorrect size information for certain domains.
7231 45. #include "config.h" was missing at the start of the crypt16.c module. This
7232 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7235 46. If there was a timeout during a "random" callout check, Exim treated it as
7236 a failure of the random address, and carried on sending RSET and the real
7237 address. If the delay was just some slowness somewhere, the response to the
7238 original RCPT would be taken as a response to RSET and so on, causing
7239 mayhem of various kinds.
7241 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7242 when I implemented it. It didn't allow for the fact that some option values
7243 may legitimately be negative (e.g. size_addition), and it didn't even do
7244 the right test for positive values.
7246 48. Domain names in DNS records are case-independent. Exim always looks them up
7247 in lower case. Some resolvers return domain names in exactly the case they
7248 appear in the zone file, that is, they may contain uppercase letters. Not
7249 all resolvers do this - some return always lower case. Exim was treating a
7250 change of case by a resolver as a change of domain, similar to a widening
7251 of a domain abbreviation. This triggered its re-routing code and so it was
7252 trying to route what was effectively the same domain again. This normally
7253 caused routing to fail (because the router wouldn't handle the domain
7254 twice). Now Exim checks for this case specially, and just changes the
7255 casing of the domain that it ultimately uses when it transmits the message
7258 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7261 50. If a filter generated a file delivery with a non-absolute name (possible if
7262 no home directory exists for the router), the forbid_file option was not
7265 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7266 the existing equality matching.
7268 52. Exim was using ints instead of ino_t variables in some places where it was
7269 dealing with inode numbers.
7271 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7272 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7273 called TMPDIR, and if it finds it is different, it changes its value.
7275 54. The smtp_printf() function is now made available to local_scan() so
7276 additional output lines can be written before returning. There is also an
7277 smtp_fflush() function to enable the detection of a dropped connection.
7278 The variables smtp_input and smtp_batched_input are exported to
7281 55. Changed the default runtime configuration: the message "Unknown user"
7282 has been removed from the ACL, and instead placed on the localuser router,
7283 using the cannot_route_message feature. This means that any verification
7284 failures that generate their own messages won't get overridden. Similarly,
7285 the "Unrouteable address" message that was in the ACL for unverifiable
7286 relay addresses has also been removed.
7288 56. Added hosts_avoid_esmtp to the smtp transport.
7290 57. The exicyclog script was not checking for the esoteric option
7291 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7292 will work only if exicyclog is run under the appropriate euid.
7294 58. Following a discussion on the list, the rules by which Exim recognises line
7295 endings on incoming messages have been changed. The -dropcr and drop_cr
7296 options are now no-ops, retained only for backwards compatibility. The
7297 following line terminators are recognized: LF CRLF CR. However, special
7298 processing applies to CR:
7300 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7301 nor a local message in the state where . is a terminator.
7303 (ii) If a bare CR is encountered in a header line, an extra space is added
7304 after the line terminator so as not to end the header. The reasoning
7305 behind this is that bare CRs in header lines are most likely either
7306 to be mistakes, or people trying to play silly games.
7308 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7309 was being incorrectly given as 18 bytes larger than it should have been.
7310 This is a VOB (very old bug).
7312 60. This may never have affected anything current, but just in case it has:
7313 When the local host is found other than at the start of a list of hosts,
7314 the local host, those with the same MX, and any that follow, are discarded.
7315 When the list in question was part of a longer list of hosts, the following
7316 hosts (not currently being processed) were also being discarded. This no
7317 longer happens. I'm not sure if this situation could ever has previously
7320 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7323 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7324 header. This is something that is recommended in a new Internet Draft, and
7325 is something that is documented as being done by Sendmail. There are two
7326 possible values. For messages generated by the autoreply transport, Exim
7329 Auto-Submitted: auto-replied
7331 whereas for all other generated messages (e.g. bounces) it adds
7333 Auto-Submitted: auto-generated
7335 63. The "personal" condition in filters now includes a test for the
7336 Auto-Submitted: header. If it contains the string "auto-" the message it
7337 not considered personal.
7339 64. Added rcpt_include_affixes as a generic transport option.
7341 65. Added queue_only_override (default true).
7343 66. Added the syslog_duplication option.
7345 67. If what should have been the first header line of a message consisted of
7346 a space followed by a colon, Exim was mis-interpreting it as a header line.
7347 It isn't of course - it is syntactically invalid and should therefore be
7348 treated as the start of the message body. The misbehaviour could have
7349 caused a number of strange effects, including loss of data in subsequent
7350 header lines, and spool format errors.
7352 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7353 client host had authenticated. This control can now be exercised by an ACL
7354 for more flexibility.
7356 69. By default, callouts do not happen when testing with -bh. There is now a
7357 variant, -bhc, which does actually run the callout code, including
7358 consulting and updating the callout cache.
7360 70. Added support for saslauthd authentication, courtesy of Alexander
7363 71. If statvfs() failed on the spool or log directories while checking their
7364 size for availability, Exim confusingly gave the error "space shortage".
7365 Furthermore, in debugging mode it crashed with a floating point exception.
7366 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7367 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7368 serious problem, Exim now writes to the main and panic logs when this
7369 happens, with details of the failure. It then refuses to accept the
7370 incoming message, giving the message "spool directory problem" or "log
7371 directory problem" with a 421 code for SMTP messages.
7373 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7374 0, 1, and 2 exist, because some OS complain for execs without them (see
7375 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7376 descriptors. However, the code omitted to check that the open succeeded,
7377 causing mysterious errors if for some reason the permissions on /dev/null
7378 got screwed. Now Exim writes a message to the main and panic logs, and
7379 bombs out if it can't open /dev/null.
7381 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7382 interact so that it is all more flexible. It is supposed to remain
7383 backwards compatible. Also added extra_local_interfaces.
7385 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7386 to bomb out with an assertion failure - to the client this appears as a
7387 connection drop. This problem occurs in the part of the code that was taken
7388 from the Samba project. Fortunately, the assertion is in a very simple
7389 function, so I have fixed this by reproducing the function inline in the
7390 one place where it is called, and arranging for authentication to fail
7391 instead of killing the process with assert().
7393 75. The SPA client code was not working when the server requested OEM rather
7394 than Unicode encoding.
7396 76. Added code to make require_files with a specific uid setting more usable in
7397 the case where statting the file as root fails - usually a non-root-mounted
7398 NFS file system. When this happens and the failure is EACCES, Exim now
7399 forks a subprocess and does the per-uid checking as the relevant uid.
7401 77. Added process_log_path.
7403 78. If log_file_path was not explicitly set, a setting of check_log_space or
7404 check_log_inodes was ignored.
7406 79. If a space check for the spool or log partitions fails, the incident is now
7407 logged. Of course, in the latter case the data may get lost...
7409 80. Added the %p formatting code to string_format() so that it can be used to
7410 print addresses in debug_print(). Adjusted all the address printing in the
7411 debugging in store.c to use %p rather than %d.
7413 81. There was a concern that a line of code in smtp_in.c could overflow a
7414 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7415 initially expressed, the concern was not well-founded, because trailing
7416 spaces are removed early. However, if the trailing spaces were followed by
7417 a NULL, they did not get removed, so the overflow was possible. Two fixes
7420 (a) I re-wrote the offending code in a cleaner fashion.
7421 (b) If an incoming SMTP command contains a NULL character, it is rejected
7424 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7425 runs initgroups(), so that if the Exim user is in any additional groups,
7426 they will be used during message reception.
7432 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob